To facilitate that this change removes it form the default filter
list. By default nova has used placement for AZs so this filter
has not been requried since xena.
Change-Id: Ie5e216dd8c2a7ecf43cc6954ec4f73d4d67b5b3b
(cherry picked from commit ad029c0e8b)
(cherry picked from commit 0726a4f919)
This patch: https://review.opendev.org/c/openstack/devstack/+/882299
provides functionality, that commit hash can be passed as last arugment,
however when GIT_DEPTH is set, it fails, as in:
timeout -s SIGINT 0 git clone https://github.com/ovn-org/ovn.git ./ovn
--depth 1 --branch 03b95a4566a15f7544f4cdf35629dacede4dcf55
fatal: Remote branch 03b95a4566a15f7544f4cdf35629dacede4dcf55
not found in upstream origin
Closes-Bug: #2023020
Change-Id: I748354964a133e028e12458cc9014d6d014cbdb9
Instead of using RDO Trunk repo server, CentOS official mirrors provide
a most reliable infrastructure and supports EMS which is required when
enabling FIPS in C9S.
In order to install the rdo-release rpm from repo.fedoraproject.org,
which does not support EMS, I'm using a workaround to wget, which works
with non-EMS servers because it uses gnutls instead of openssl, and
install it locally with rpm.
This is also consistent to CentOS 8 implementatioin.
Closes-Bug: #2020661
Closes-Bug: #2020434
Change-Id: Icd99f467d47aaafaaf3ee8f2a3c4da08842cb672
(cherry picked from commit b2ad00cb66)
This increases the timeout we use to wait for cinder to perform a
volume reimage. Since devstack is often running on a single machine
with non-production IO performance, we should bump this limit to avoid
hitting it before the rebuild completes.
Change-Id: Ie2663b951acb0c1a65597a39e032948764e6ae6a
(cherry picked from commit 58c80b2424)
(cherry picked from commit 6767465bc5)
git_clone assumes a branch or a tag is passed as the last argument, and
it fails when a commit hash is passed, as in:
timeout -s SIGINT 0 git clone https://github.com/ovn-org/ovn.git
/opt/stack/ovn --branch 36e3ab9b47e93af0599a818e9d6b2930e49473f0
Cloning into '/opt/stack/ovn'...
fatal: Remote branch 36e3ab9b47e93af0599a818e9d6b2930e49473f0 not found
in upstream origin
Change-Id: Id1328d7cba418fa7c227ae9db4fe83c09fd06035
(cherry picked from commit e8915786e1)
Irrespective of build_modules is True
or False reload ovs modules always.
If ovs is installed from package before(like
with multi-node-bridge role), then installing
ovs from source requires openvswitch kernel
module to be reloaded.
The issue was not seen before jammy as there
module was reloaded when build_modules was set
to True.
Closes-Bug: #2015364
Change-Id: I1785b49b2ef72ca1f817f504d5ea56021410c052
(cherry picked from commit 42517968ff)
(cherry picked from commit f1d555d1e7)
These are a few tweaks I applied to my own memory-constrained cloud
instances that seemed to help. I have lower performance requirements
so this may make things worse and not better, but it's worth seeing
what the impact is. I'll admit to not knowing the full impact of these
as they're mostly collected from various tutorials on lowering memory
usage.
Enable this for now on devstack-multinode
Change-Id: I7b223391d3de01e3e81b02076debd01d9d2f097c
(cherry picked from commit 7567359755)
Creating multiattach volume is a non-admin operation but creating
multiattach volume type is an admin operation.
Previously cinder allowed creating multiattach volumes without a
volume type but that support is being removed with[1].
The change requires updating tempest tests[2] but some tempest
tests are non-admin, which require admin priviledges to create the
multiattach volume type.
Based on the last discussion with tempest team[3], the proposed
solution is to create a multiattach volume type in devstack,
if ENABLE_VOLUME_MULTIATTACH is True, and use it in tempest
tests. Similar to how admins create multiattach volume types
for non-admin users.
This patch creates a multiattach volume type if
ENABLE_VOLUME_MULTIATTACH is True. Also we set the multiattach
type name as a tempest config option 'volume_type_multiattach'.
[1] https://review.opendev.org/c/openstack/cinder/+/874865
[2] https://review.opendev.org/c/openstack/tempest/+/875372
[3] https://meetings.opendev.org/irclogs/%23openstack-cinder/%23openstack-cinder.2023-03-13.log.html#t2023-03-13T18:47:56
Change-Id: Icd3690565bf7b27898cd206641e612da3993703d
(cherry picked from commit 1898a683be)
(cherry picked from commit 16594b7b2a)
Conflicts: lib/cinder
conflicts due to nvme support being added in 2023.1
https://review.opendev.org/c/openstack/devstack/+/814193
Devstack set the env var TOX_CONSTRAINTS_FILE/UPPER_CONSTRAINTS_FILE
which are used to use the constraints during Tempest virtual env installation.
Those env var are set to non-master constraint when we need to use non-master
constraints but when we need to use the master constraints we do not set/reset
them point to master constraints. This create the issue when running the grenade
job where we run Tempest on the old devstack as well as in the new devstack.
When tempest is installed on old devstack then old tempest is used and it sets
these env var to stable/<branch> constraints (this is the case when old devstack
(the stable branch is in EM phase) uses the old tempest not the master tempest),
all good till now. But the problem comes when in the same grenade script run
upgrade-tempest install the master tempest (when new devstack branches are in
the 'supported' phase and use the master tempest means) and are supposed to use
the master constraints. But the TOX_CONSTRAINTS_FILE/UPPER_CONSTRAINTS_FILE env
var set by old tempest is used by the tempest and due to a mismatch in constraints
it fails.
This happened when we tried to pin the stable/wallaby with Tempest 29.0.0
- https://review.opendev.org/c/openstack/devstack/+/871782
and table/xena grenade job failed (stable/xena use master tempest and supposed
to use master constraints)
- https://zuul.opendev.org/t/openstack/build/fb7b2a8b562c42bab4c741819f5e9732/log/controller/logs/grenade.sh_log.txt#16641
We should set/reset those constraint env var to master constraints if configuration
tell devstack to use the master constraints.
[1] https://github.com/openstack/devstack/blob/71c3c40c269a50303247855319d1d3a5d30f6773/lib/tempest#L124
Closes-Bug: #2003993
Change-Id: I5e938139b47f443a4c358415d0d4dcf6549cd085
(cherry picked from commit 7fe998109b)
Just like we remove db files let's also remove
socket files when initializing ovn. Those will
reappear once service fully restarts along with
db files. Without it we see random issue as
described in the below bug.
Closes-Bug: #2002629
Change-Id: I726a9cac9c805d017273aa79e844724f0d00cdf0
(cherry picked from commit 7fecba2f13)
Tox 4.0.0 has some incompatible changes, epecially more
strict on allowlist_externals. Tempest recently changed
allowlist_externals not to be *[1] causing the failure
on jobs where lib/tempest failing to run the tempest
as command in virtual env.
----------
venv: commands[0]> tempest verify-config -uro /tmp/tmp.qH5KgJHTF4
venv: failed with tempest is not allowed, use allowlist_externals to allow it
------
We do not need to test/fix the <=stable/zed branches with tox 4.0.0
and pinning them with the compatible tox version of the time stable
brnaches were releaased is better way.
This commit proposes:
1. Pinning the tox<4.0.0 for <=stable/ze branches testing
2. Workaround to unblock the master gate by pinning it <4.0.0 but
we should make our testing compatible with tox 4.0.0 soon.
Depends-On: https://review.opendev.org/c/openstack/devstack/+/867067
Related-Bug: #1999183
[1] https://review.opendev.org/c/openstack/tempest/+/865314 devstack based job started failing to run tempest command on venv.
Change-Id: I9a138af94dedc0d8ce5a0d519d75779415d3c30b
(cherry picked from commit ba54baa425)
These jobs are not intended to run on stable branches; remove them
to avoid issues with updating the nodeset on the master branch:
- devstack-platform-fedora-latest
- devstack-platform-fedora-latest-virt-preview
Change-Id: I4044c20765ad96fb5304d145b2c18be9736de7f3
As added in notes for INSTALL_TEMPEST variable, we need to set
this to False for a stable branch so that devstack does not
install Tempest system wide.
Change-Id: I8a79feedb676514dfe9c3125c9d8e66417547d52
This commit caps the network, volume and swift extensions on
Tempest's config option api_extensions.
Change-Id: I33d17c81a40861284087de337bc33b3d9e60f076
Update the URL to the upper-constraints file to point to the redirect
rule on releases.openstack.org so that anyone working on this branch
will switch to the correct upper-constraints list automatically when
the requirements repository branches.
Until the requirements repository has as stable/zed branch, tests will
continue to use the upper-constraints list on master.
Change-Id: I7089a04cff68145388f12dab18c6cdd1a5a96336
This patch changes user who runs ovsdb-server and ovn-nortd services
to root.
It also adds installation of the libssl dev package before compilation
of the openvswitch if TLS service is enabled.
Co-Authored-By: Fernando Royo <froyo@redhat.com>
Closes-Bug: #1987832
Change-Id: I83fc9250ae5b7c1686938a0dd25d66b40fc6c6aa
In case of online mode, there is a procedure to recreate tempest venv.
For consistency of tempest venv during the entire stack.sh process,
add logic to consider the TEMPEST_VENV_UPPER_CONSTRAINTS option here.
Closes-bug: #1980483
Signed-off-by: June Yi <june.yi@samsung.com>
Change-Id: I0cea282152fd363af8671cab1b5f733ebe2bd4df
neutron-ns-metadata-proxy was dropped from Neutron 5 years ago, no need
to keep trying to kill it.
Change-Id: I20b6d68dd8dde36057a2418bca0841bdea377b07
When barbican is enabled, add the "creator" role to cinder's service
user so that cinder can create secrets. Cinder needs to create
barbican secrets when migrating encryption keys from the legacy
ConfKeyManager to barbican. Cinder also needs to create barbican
secrets in order to support transferring encrypted volumes.
Implements: bp/transfer-encrypted-volume
Depends-On: I216f78e8a300ab3f79bbcbb38110adf2bbec2196
Change-Id: Ia3f414c4b9b0829f60841a6dd63c97a893fdde4d
In I90316208d1af42c1659d3bee386f95e38aaf2c56 support for nova-network
was removed, but some bits remained, fix this up.
Change-Id: Iba7e1785fd0bdf0a6e94e5e03438fc7634621e49
Openstack client can return the id field for create/show commands using
`-f value -c id`. Cleaned up the use of grep 'id' with get_field
Change-Id: I2f4338f30c11e5139cda51c92524782b86f0aacc
Right now we don't officialy support LinuxMint as our
documentation says [1], it seems LinuxMint is a relict
and got forgotten over time.
This patch removes LinuxMint from the code in order not to
confuse users.
[1] https://docs.openstack.org/devstack/latest/
Closes-Bug: #1983427
Change-Id: Ie1ced25f89389494b28a7b2e9bb1c4273e002dd5
Attempting to set a value containing the ampersand
character (&) by iniset would corrupt the value.
So, add an escaping process.
Signed-off-by: Nobuhiro MIKI <nmiki@yahoo-corp.jp>
Closes-Bug: #1983816
Change-Id: Ie2633bacd2d761d110e6cb12f95382325c329415
Writing NOPASSWD directive into /etc/sudoers was throwing
permission denied errors. This commit writes the directive
to the /etc/sudoers.d/stack file instead.
Closes-Bug: #1981541
Change-Id: If30f01aa5f3a33dda79ff4a6892116511c8e1542
Recently the experimental mechanism has been added to Neutron and now
it requires the [experimental] linuxbridge option when the linuxbridge
mechanism driver is used.
Depends-on: https://review.opendev.org/c/openstack/neutron/+/845181
Change-Id: Ice82a391cda9eb0193f23e6794be7ab3df12c40b
We see some cases where OVN startup takes much longer than 5 seconds, up
to 28 seconds have been observed, so increase the limit to 40 to be on
the safe side.
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Closes-Bug: 1980421
Change-Id: I6da4a537e6a8d527ff71a821f07164fc7d342882
Because of the missing "$" before ENFORCE_SCOPE in the lib/neutron
module, it was treated as an ENFORCE_SCOPE string instead of variable
and Neutron was deployed always with old defaults and disabled scope
enforcement.
Change-Id: Ibe67fea634c5f7abb521c0369ff30dd5db84db8c
Some of the API services are not properly mounted under /$service/
in the apache proxy. This patch tries to avoid recording data
for "services" like "v2.0" (in the case of neutron) by only adding
names if they're all letters. A single warning is emitted for any
services excluded by this check.
For the moment this will mean we don't collect data for those services,
but when their devstack API config is fixed, they'll start to show
up.
Change-Id: I41cc300e89a4f97a008a8ba97c91f0980f9b9c3f
Currently Devstack uses short hostname for configuration of OVN.
This leads to inability to start instances (failing port binding)
on hosts with full hostnames (including dots). Open vSwitch expects
hostname in external_ids that corresponds to one returned by
``hostname`` command.
Closes-Bug: #1943631
Change-Id: I15b71a49c482be0c8f15ad834e29ea1b33307c86
Due to the below bug the job has been constantly failing.
Let's make it n-v until the bug is resolved:
- https://bugs.launchpad.net/neutron/+bug/1979047
Change-Id: Ifc8cc96843a8eac5c98cd1e1f9e4b6287a7f2e7c
This patch reduces memory usage on the Cinder Volume and Backup services
by tuning glibc.
The specific tuning consist on disabling the per thread arenas and
disabling dynamic thresholds.
The Cinder Backup service suffers from high water mark memory usage and
uses excessive memory. As an example just after 10 restore operations
the service uses almost 1GB of RAM and does not ever free it afterwards.
With this patch the memory consumption of the service is reduced down to
almost 130MB. If we add a revert from Cinder (Change-Id
I43a20c8687f12bc52b014611cc6977c4c3ca212c) it goes down to 100MB during
my tests.
This glibc tuning is not applied to all Python services because I
haven't done proper testings on them and at first glance they don't seem
to have such great improvements.
Related-bug: #1908805
Change-Id: Ic9030d01468b3189350f83b04a8d1d346c489d3c
We are seeing failures when using an updated setuptools version
installed together with distro pip on Ubuntu 22.04. Install the version
from u-c only when we are also installing pip from upstream.
Change-Id: Ibb6e9424e5794ccbf9a937d2eecfa3bf60ed312e
Currently, neutron tunnel endpoints must be IPv4 addresses,
i.e. $HOST_IP, although IPv6 endpoints are supported by most
drivers.
Create a TUNNEL_IP_VERSION variable to choose which host IP
to use, either HOST_IP or HOST_IPV6, and configure it in the
OVS and Linuxbridge agent driver files. The default is still
IPv4, but it can be over-ridden by specifying TUNNEL_ENDPOINT_IP
accordingly.
This behaves similar to the SERVICE_IP_VERSION option, which
can either be set to 4 or 6, but not 4+6 - the tunnel overhead
should be consistent on all systems in order not to have MTU
issues.
Must set the ML2 overlay_ip_version config option to match
else agent tunnel sync RPC will not work.
Must set the OVN external_ids:ovn-encap-ip config option to
the correct address.
Updated 'devstack-ipv6-only' job definition and verification role
that will set all services and tunnels to use IPv6 addresses.
Closes-bug: #1619476
Change-Id: I6034278dfc17b55d7863bc4db541bbdaa983a686
When cephadm is used, if ENABLE_CEPH_C_BAK is True both pool and
key are created by devstack-plugin-ceph. This piece of code can
still stay here to make sure the cinder config is properly built.
Change-Id: I799521f008123b8e42b2021c1c11d374b834bec3
Some of the hardest-to-debug issues are qemu crashes deep in a nova
workflow that can't be reproduced locally. This adds a post task to
the playbook so that we capture the most recent qemu core dump, if
there is one.
Change-Id: I48a2ea883325ca920b7e7909edad53a9832fb319
Swift removed sha1 from supported digests with [1] and
that broked ironic tinyipa job. Temorary add sha1 to
allowed_digests until it's fixed in ironic.
[1] https://review.opendev.org/c/openstack/swift/+/525771
Story: 2010068
Task: 45539
Change-Id: I68dfc472ce901058b6a7d691c98ed1641d431e54
Grenade jobs stop services, check fip connectivity
for a nova server and then upgrade to next release.
But since ovn data plane and db services are stopped along
with other services, fip connectivity fails as a result.
We shouldn't stop these services along with other
neutron services. This patch adds a new variable
"SKIP_STOP_OVN" which can be used by grenade jobs
to skip stop of ovn services.
This will also fix the ovn grenade jobs.
Also source fixup_stuff.sh so function fixup_ovn_centos
is available. It's already sourced in stack.sh but
that's not used in grenade run.
Change-Id: I94818a19f19973779cb2e11753d2881d54dfa3bc
Packages for OVN are now available in bullseye, so we can drop the
special handling.
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I5e5c78aa19c5208c207ddcf14e208bae8fbc3c55
RHEL based distros set homedir permissions to 700,
and Ubuntu 21.04+ to 750[1], i.e missing executable
permission for group or others, this results into failures
as defined in the below bug.
Since in doc we add useradd command, it's good to
add instructions to fix the permissions there itself
instead of getting failures during installation and then
fixing it.
Also update user create script to fix permissions
by adding executable bit to DEST directory if missing.
[1] https://discourse.ubuntu.com/t/private-home-directories-for-ubuntu-21-04-onwards/19533
Closes-Bug: #1966858
Change-Id: Id2787886433281238eb95ee11a75eddeef514293
This can happen if devstack fails to run, but we still run the post
tasks. Also could happen if some sort of hybrid job configuration
does not run all of devstack but we still end up running post jobs.
Just warn to stderr and assume no DB info.
Change-Id: I211a331ab668dbb0ad7882908cca4363f865d924
The previous change, I237f5663b0f8b060f6df130de04e17e2b1695f8a, removed
a SETUPTOOLS flag, but not the comment explaining why that flag was
previously set. Clean up that comment.
Change-Id: I32b0240fd56310d7f10596aaa8ef432679bfd66a
There are two problems with dbcounter installation on Jammy. The first
is straightforward. We have to use `py_modules` instead of `modules` to
specify the source file. I don't know how this works on other distros
but the docs [0] seem to clearly indicate py_modules does this.
The second issue is quite an issue and requires story time. When
pip/setuptools insteall editable installs (as is done for many of the
openstack projects) it creates an easy-install.pth file that tells the
python interpreter to add the source dirs of those repos to the python
path. Normally these paths are appended to your sys.path. Pip's isolated
build env relies on the assumption that these paths are appeneded to the
path when it santizes sys.path to create the isolated environemnt.
However, when SETUPTOOLS_SYS_PATH_TECHNIQUE is set to rewrite the paths
are not appended and are inserted in the middle. This breaks pip's
isolated build env which broke dbcounter installations. We fix this by
not setting SETUPTOOLS_SYS_PATH_TECHNIQUE to rewrite. Upstream indicates
the reason we set this half a decade ago has since been fixed properly.
The reason Jammy and nothing else breaks is that python3.10 is the first
python version to use pip's isolated build envs by default.
I've locally fiddled with a patch to pip [1] to try and fix this
behavior even when rewrite is set. I don't plan to push this upstream
but it helps to illustrate where the problem lies. If someone else would
like to upstream this feel free.
Finally this change makes the jammy platform job voting again and adds
it to the gate to ensure we don't regress again.
[0] https://docs.python.org/3/distutils/sourcedist.html#specifying-the-files-to-distribute
[1] https://paste.opendev.org/show/bqVAuhgMtVtfYupZK5J6/
Change-Id: I237f5663b0f8b060f6df130de04e17e2b1695f8a
The job is broken since it is running with python3.7 and most services
now require at least python3.8.
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: Ie21f71acffabd78c79e2b141951ccf30a5c06445
We missed to add the jobs to the gate queue and so they have already
regressed before they were actually in place. Make them non-voting for
now until the issues are fixed.
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I5d1f83dfe23747096163076dcf80750585c0260e
Glance image import is asynchronous and may be configured to do image
conversion. If image import is being used, it's possible that the
tempest configuration code is executed before the import has
completed and there may be no active images yet. In that case,
we will poll glance every TEMPEST_GLANCE_IMPORT_POLL_INTERVAL seconds
(default: 1) to see if there are TEMPEST_GLANCE_IMAGE_COUNT active
images (default: 1) up to TEMPEST_GLANCE_IMPORT_POLL_LIMIT times
(default: 12).
You can see an example of the issue this patch addresses in real
life:
https://review.opendev.org/c/openstack/glance/+/841278/1#message-456096e48b28e5b866deb8bf53e9258ee08219a0
Change-Id: Ie99f12691d9062611a8930accfa14d9540970cc5
Without it segment plugin fails to connect with
placement api. Configure the placement section
if service is deployed.
Closes-Bug: #1973783
Change-Id: Ie7f37770a04f622735cf2263c601257669ab5064
Two runs of the same job on the same patch can yield quite different
numbers for API calls if we just count the raw calls. Many of these
are tempest polling for resources, which on a slow worker can require
many more calls than a fast one.
Tempest seems to not change its User-Agent string, but the client
libraries do. So, if we ignore the regular "python-urllib" agent
calls, we get a much more stable count of service-to-service API
calls in the performance report.
Note that we were also logging in a different (less-rich) format for
the tls-proxy.log file, which hampers our ability to parse that
data in the same format. This switches it to "combined" which is used
by the access.log and contains more useful information, like the
user-agent, among other things.
Change-Id: I8889c2e53f85c41150e1245dcbe2a79bac702aad
The mysql performance_schema method for counting per-database queries
is very heavyweight in that it requires full logging (in a table) of
every query. We do hundreds of thousands in the course of a tempest
run, which ends up creating its own performance problem.
This changes the approach we take, which is to bundle a very tiny
sqlalchemy plugin module which counts just what we care about in
a special database.
It is more complex than just enabling the features in mysql, but it
is a massively smaller runtime overhead. It also provides us the
opportunity to easily zero the counters just before a tempest run.
Change-Id: I361bc30bb970cdaf18b966951f217862d302f0b9
The new Ubuntu LTS release has been made last week, start running
devstack on it as a platform job.
Horizon has issues with py310, so gets disabled for now.
Run variants with OVS and OVN(default).
Co-Authored-By: yatinkarel <ykarel@redhat.com>
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I47696273d6b009f754335b44ef3356b4f5115cd8
When OVN is setup from distro packages, the
main service is ovn-central which when restarted,
restarts ovn-northd, ovn nb and db services.
And during the restart ovn dbs(ovnnb_db.db and ovnsb_db.db)
are created, which may sometime takes time as seen with
ubuntu jammy tests[1].
We already checking for socket's file to be available,
let's also check for db files as without it ovn-*ctl
operations succeed but changes are not persisted until
db files are available and changes are lost with the restart.
[1] https://review.opendev.org/c/openstack/devstack/+/839389
Change-Id: I178da7af8cba8bcc8a67174e439df7c0f2c7d4d5
Would be helpful in troubleshooting services
which either fails to start or takes time to
start.
Related-Bug: #1970679
Change-Id: Iba2fce5f8b1cd00708f092e6eb5a1fbd96e97da0
In order to run on systems where not all requirements are present,
we should be tolerant of missing external dependencies, such as
psutil and pymysql. Print a warning (to stderr) and just leave out
those stats in that case.
Also make running the stats collector use ignore_errors:yes to avoid
failures in the future. I think the stats is not critical enough to
fail a job for bugs like this.
Related-Bug: #1970195
Change-Id: I132b0e1f5033c4f109a8b8cc776c0877574c4a49
In some cases the value is [not set], in this case
the conversion to integer does not work.
Closes-Bug: #1970431
Change-Id: I74df7d8bc9f5cbe0709a6471cf7639caea0b58e8
A long time ago, Ironic's IPv6 only job started to fail working with
errors indicated the host was unreacable. Turns out, this was because
the $ext_gw_interface was not being set to up, and thus could
be found in a Down state, and thus the kernel would not accept routes
for it.
Adds an explicit step to turn up the public bridge, much as done in
the IPv4 router plugin code which would also be executed in 4+6.
That being said, Ironic's CI jobs are very intentionally IPv6 only
to ensure that we have no chances of v4 addressing getting used
at any point in time.
This should allow Ironic to return it's IPv6 only CI job back
to the normal check queue, once a ironic plugin issue has been
resolved which was introduced while it was removed.
Change-Id: I121ec8a2e9640b21a7126f2eeb23da36b4aa95bf
This updates each devstack service library, to use it as the
default value for service-specific RBAC configuration.
Change-Id: I41061d042206c411ee3dd94ce91098e612af7ae7
I941ef5ea90970a0901236afe81c551aaf24ac1d8 added a sed command that
should match and delete path values but used '/' as sed separator. This
leads to error in unstack.sh runs when the path also contains '/':
+./unstack.sh:main:188 sudo sed -i '/directory=/opt/stack/ d' /etc/gitconfig
sed: -e expression #1, char 13: unknown command: `o'
So this patch replace '/' separator with '+'.
Change-Id: I06811c0d9ee7ecddf84ef1c6dd6cff5129dbf4b1
This change adds mod_ssl to the default set of rpms installed
on rpm based distros.
this is required if the tls-proxy service is enabled
for multi node centos based jobs.
Change-Id: I52652de88352094c824da68e5baf7db4c17cb027
the value of LOGDAYS in samples/local.conf is 2, so change the
value in the comment and the sample value in the document to
be consistent with it.
Change-Id: I5822bbf1d6ad347c67c886be1e3325113d079114
This makes us gather a bunch of consistent statistics after we run
tempest that can be use to measure the impact of a given change. These
are stable metrics such as "number of DB queries made" and "how much
memory is each service using after a tempest run."
Note that this will always run after devstack to generate the JSON
file, but there are two things that control its completeness:
- MYSQL_GATHER_PERFORMANCE must be enabled to get per-db stats
- Unless tls-proxy is enabled, we will only get API stats for keystone
Change-Id: Ie3b1504256dc1c9c6b59634e86fa98494bcb07b1
This is necessary for more consistent behavior across multiple
distro versions. Apparently somewhere along the way, git started
looking at the current user's home directory instead of $HOME.
Related-Bug: https://bugs.launchpad.net/devstack/+bug/1968798
Change-Id: I941ef5ea90970a0901236afe81c551aaf24ac1d8
git commit [1] introduced a new behaviour to work around a CVE that
disallows any git operations in directories not owned by the current
user.
This may seem unrelated to installation, but it plays havoc with PBR,
which calls out to git to get to get revision history. So if you are
"pip install"-ing from a source tree you don't own, the PBR git calls
in that tree now fail and the install blows up.
This plays havoc with our model. Firstly, we checkout all code as
"stack" then install it globally with "sudo" (i.e. root) -- which
breaks. We also have cases of essentially the opposite -- checkouts
we have installed as root, but then run tox in them as a regular user;
tox wants to install the source in its venv but now we have another
user conflict.
This uses the only available configuration option to avoid that by
globally setting the source directories we clone as safe. This is an
encroachment of the global system for sure, but is about the only
switch available at the moment. For discussion of other approaches,
see [2].
Related-Bug: https://bugs.launchpad.net/devstack/+bug/1968798
[1] https://github.com/git/git/commit/8959555cee7ec045958f9b6dd62e541affb7e7d9
[2] https://review.opendev.org/c/openstack/devstack/+/837636
Change-Id: Ib9896a99b6d6c4d359ee412743ce30512b3c4fb7
osc is typicaly installed in /usr/local/bin
to avoid command not found errors when invoking osc
in devstack ensure that /usr/local/bin is included
in the PATH.
Change-Id: I605fbc4b131149bf5d1b6307b360fe365c680b1a
OpenEuler job fails 100% of the time. As discussed in QA meeting,
we agreed to move OpenEuler job to experimental pipeline.
- https://meetings.opendev.org/meetings/qa/2022/qa.2022-03-22-15.00.log.html#l-76
Once it is fixed, we can think of adding back to regular pipeline.
Change-Id: I831889a09fabe5bed5522d17e352ec8009eac321
devstack isn't a python project, these were introduced only for docs
building and made redundant with [0]. We can remove them now.
[0] Iedcc008b170821aa74acefc02ec6a243a0dc307c
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I90ca1c6918c016d10c579fbae49d13fff1ed59af
After patch [1] project_id in that module is no longer needed as to make
it working with new secure RBAC policies we had to hardcode "demo"
project to be used always.
This is small follow-up patch with cleaning after [1].
[1] https://review.opendev.org/c/openstack/devstack/+/826851/
Change-Id: Iddf9692817c91807fc3269547910e4f83585f07f
There are a couple of places that still use a hardcoded
127.0.0.1 value, even if devstack is run with
SERVICE_IP_VERSION=6 in local.conf. While things still
work, SERVICE_LOCAL_HOST should be used instead since
everything else could be using IPv6.
Change-Id: I2dd9247a4ac19f565d4d5ecb2e1490501fda8bca
Newer versions of openssl (CentOS9Stream for example) do not like using sha1.
Devstack will fail on these systems[1] with the following error:
801B93DCE77F0000:error:03000098:digital envelope routines:do_sigver_init:invalid digest:crypto/evp/m_sigver.c:333:
This patch updates the tls-proxy code in devstack to use sha256 instead of sha1 which allows devstack to complete when tls-proxy is enabled.
[1] https://zuul.opendev.org/t/openstack/build/1d90b22a39c74e24a8390861b3c5f957/log/job-output.txt#5535
Closes-Bug: #1962600
Change-Id: I71e1371affe32f070167037b0109a489d196bd31
If the ci images do not have any cached data
we should ignore any error when trying to copying it.
This is requried when using unmodified cloud images.
Change-Id: Ia6e94fc01343d0c292b1477905f8a96a6b43bcf8
This is a followup for change Ifdef3510bc7da3098a71739814e35dbaf612ae34
which added configuration of unified limits for nova. This removes an
unnecessary wrapper unsetting of OS_ env variables, unnecessary quoting
on an iniset config value, and a hardcoding of user domain. The glance
code from which the nova code was originally copied is also cleaned up.
Change-Id: I4921af5cc0f624dd5aa848533f7049ee816be593
Previously, Swift's backing disk were not be mounted after reboots,
causing swift-proxy-server service to fail with cryptic error
messages such as 'proxy-server: ERROR Insufficient Storage'. Now,
we use Dan Smith' create_disk function from functions to create
the backing disk for us and add it to /etc/fstab.
Change-Id: I9cbccc87bc94a55b58e9badf3fdb127d6f1cf599
The rsync debian package is required for swift service. This
requirement has been covered by rpms but not for deb packages.
Change-Id: Iefd1302be9c7fd80e037bbae3638602d6d823580
This change use uname -m to get the portable host arch and uses that
as a new default. on x86_64 hosts this should result in no visable change
in behavior however on a non x86 host it will cause devstack to attempt
to download a cirros image that matches the host.
Change-Id: I6d1495a23400ef4cf496302028324fa5794dd45f
This patch fixes several issues related to the installation with
OVN backend with the OVS/OVN compilation enabled.
The OVS/OVN local directories prefix, when both services are compiled,
is now "/usr/local".
The "ovn_agent._run_process" function is calling "ovs-appctl" to
configure the logging settings of several services. Instead of
using the service name, the ctl socket file is used instead. That
is more robust and does not fail in systems with previous
installations.
Closes-Bug: #1960514
Change-Id: I69de5333393957593db6e05495f0c3c758efefdf
Previously, loop devices for LVM volume groups backing files were not
created after reboots, causing e.g. Cinder to fail with messages such
as
ERROR cinder.service [-] Manager for service cinder-volume
devstack@lvmdriver-1 is reporting problems, not sending
heartbeat. Service will appear "down".
Now, we use systemd services to manage loop devices for backing files.
Change-Id: I27ec027834966e44aa9a99999358f5b4debc43e0
This change reverts the execute permissions from
stackrc which is not meant to be run as a script but sourced
as part of stack.sh
Change-Id: I9a05051e5a297cfaf78d097fa5f90a7c5fd254a6
Since pip v22, python3.6 is not supported (the minimum version is
python3.7). This patch adds the reference for the pip3.6 URL to be
used instead of the default one.
Closes-Bug: #1959600
Change-Id: Iab2c391d5388461fe9e9037cee81884ce8032e72
This enables the configuration of nova to use unified limits in
keystone and enforcement in oslo.limit.
Related to blueprint unified-limits-nova
Depends-On: https://review.opendev.org/c/openstack/nova/+/715271
Change-Id: Ifdef3510bc7da3098a71739814e35dbaf612ae34
This reverts commit be7b5bf671.
As related bug is fixed, lets enabled scope enforcement in Neutron
again.
Related-bug: #1959196
Change-Id: I72db7ef533e78a10734d105e6a0debef288e41a1
After patch [1] new RBAC policies changed in the way that SYSTEM_ADMIN
user isn't anymore allowed to e.g. create resources in behalf of some
projects. Now PROJECT_ADMIN needs to create such resources instead.
So this patch basically reverts most of the changes which were done
in [2] some time ago.
It also introduces new entry in the clouds.yaml file -
"devstack-admin-demo" which is "admin" user in the "demo" project as
it's needed to create some resouces in the demo project now.
Additionally, because of bug [3] this patch changes way how IPv6
external gateway IP is found using Neutron API. This change may be
reverted in the future when bug [3] will be fixed.
[1] https://review.opendev.org/c/openstack/neutron/+/821208
[2] https://review.opendev.org/c/openstack/devstack/+/797450
[3] https://bugs.launchpad.net/neutron/+bug/1959332
Depends-On: https://review.opendev.org/c/openstack/neutron/+/826828
Closes-Bug: #1959196
Change-Id: I32a6e8b9b59269a8699644b563657363425f7174
After patch [1] was merged in Neutron, enforcing scopes there
is broken.
So lets disable it temporary to unblock Devstack's gate for now.
[1] https://review.opendev.org/c/openstack/neutron/+/821208
Related-Bug: #1959196
Change-Id: I24da6f3897a638749d16f738329a873a5f9a291d
This reverts commit 26bd94b45e.
Reason for revert: Devstack keystone creation/setup are moved to
scope tokens, so we can reintroduce the scope check enable.
Change-Id: I6e1c261196dbcaf632748fb6f04e0867648b76c7
This is needed so we can set keystone into enforcing secure RBAC.
This also adjusts lib/glance, which already partially used
devstack-system-admin.
Change-Id: I6df8ad23a3077a8420340167a748ae23ad094962
Running get-pip.py fails on Ubuntu when running twice, e.g. after a
unstack/stack cycle. Just use distro pip instead.
Closes-Bug: #1957048
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I87a8d53ed8860dd017a6c826dee6b6f4baef3c96
The default Neutron configuration is now using OVN, but the multinode
lab was using an incompatible configuration:
The q-agt/neutron-agt service must be disabled with OVN.
Change-Id: I518a739a3daac941880463cde6b47951331d0911
Recent iputils release in CentOS 8-stream causing
ping failures with non root user. This needs a fix
in systemd package as mentioned in the Related Bugs,
until it's fixed and is in 8-stream mirrors let's
workaround it by setting net.ipv4.ping_group_range
setting manually.
Related-Bug: #1957941
Related-Bug: rhbz#2037807
Change-Id: I0d8dac910647968b625020c2a94e626ba5255058
This patch allows to skip the installation
of the database backend packages (MySQL or Postgres)
with the introduction of the INSTALL_DATABASE_SERVER_PACKAGES
variable (defaulted to True).
This is useful in such environments that do not require
to install the MySQL/Postgres server packages directly but using
a container serving that purpose, for those cases all the
remaining steps should be executed just skipping the
packages install.
Change-Id: I26628a31fdda3ce95ed04a2b7ae7b132c288581f
We always need the master branch of requirements in order to be able to
install tempest with it, so override GIT_DEPTH when cloning that repo.
Closes-Bug: 1956616
Change-Id: Id0b409bfadd73f2c30314724178d6e199121050b
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
For some reason, setting the CHAPAlgorithms as in c3b705138
breaks OpenEuler. Making this conditional so that tests continue
to pass.
Change-Id: Iaa740ecfbb9173dd97e90485bad88225caedb523
There was no space after the --project option in the command that
creates the public subnet, thus if any option follows, the option itself
will be parsed as part of the value passed to the --project option. This
change just adds the missing space.
Change-Id: I1e7375578342a82717222e902fcd65a4a62e33a7
When deploying devstack in a stable branch, the master branch is
available locally only in a CI environment where Zuul prepares all
available branches. For a non-CI deployment we need to stick to using
the remote branch as was the case before [0].
While the situation on the master branch isn't really broken, we apply
the fix here anyway so that future stable branches are created in a
working state.
[0] I5d42ac6b54bf20804d7e5faa39d1289102318b64
Closes-Bug: #1956219
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: Ib7719cb2d48b34db70f885e0afe77d904abba3b5
mysqladmin is incorrectly installed in Fedora 34 with mariadb. This
causes the failure of Zuul Fedora based jobs. The issue is a conflict
between mariadb and community mysql that is described in [1] and [2].
The workaround is to explicitly install package "mariadb"
Also configure an increased swap size like for the other platform jobs
in order to avoid OOM issues.
[1] https://bugzilla.redhat.com/show_bug.cgi?id=2026933
[2] https://lists.launchpad.net/maria-discuss/msg06179.html
Closes-Bug: #1956116
Change-Id: Icf6d7e1af5130689ea10b29d37cc9b188b2c9754
We still are seeing regular job failures because the pcp package fails
to install. Assume that we can still enable it on demand when someone
needs to debug specific job issues, let us just disable it by default.
Related-Bug: 1943184
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I32ef8038e21c818623db9389588b3c6d3f98dcad
When we need to read a DATABASE_PASSWORD from the user, make sure we
actually use it in our database URLs.
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I5ebf6b0280e82f2c87a63cbee7a9957c6bd26898
That function was accepting 3 positional arguments and first
of them was boolean value "build_modules" which isn't used anywhere in
that function.
So this patch cleans it a bit by removing that not used parameter.
Change-Id: I5c57b9116338a63b7bfb170c02e33bb4eae725da
openEuler is an open-source Linux based operating system. The current
openEuler kernel is based on Linux and supports multi arch, such as X86_64
and aarch64. It fully unleashes the potential of computing chips. As an
efficient, stable, and secure open-source OS built by global open-source
contributors, openEuler applies to database, big data, cloud computing,
and AI scenarios. openEuler is using RPM for package management.
Note:
Currently there is no available package for uwsgi-plugin-python3 and ovn, so that
openEuler needs manually install them from source.
Website: https://www.openeuler.org/en/
Change-Id: I169a0017998054604a63ac6c177d0f43f8a32ba6
Co-Authored-By: wangxiyuan <wangxiyuan1007@gmail.com>
Signed-off-by: Kevin Zhao <kevin.zhao@linaro.org>
This patch adds new config option NEUTRON_ENFORCE_NEW_DEFAULTS which
if set to True will deploy Neutron with enforce new rbac defaults and
scopes.
It will also use SYSTEM_ADMIN user to interact with Neutron where it is
needed.
Depends-On: https://review.opendev.org/c/openstack/neutron/+/798821
Change-Id: I14d934f0deced34d74003b92824cad3c44ec4f5e
The default CHAP algorithm for iscsid is md5, which is disallowed
under fips. We will set the chap algorithm to "SHA3-256,SHA256",
which should work under all configurations.
Change-Id: Ide186fb53b3f9826ff602cb7fb797f245a15033a
Function _neutron_ovs_base_install_agent_packages always tried to
install openvswitch from packages and start it using systemd units.
That was failing when ovs was expected to be installed from source.
This patch fixes that.
Change-Id: Iae8625dd800d30061ea3dbed9eb0dfbe16f21572
Previously this would always happen for Nova and Cinder even if n-api
and c-api were not enabled on the host respectively.
This change stops this by placing both calls write_uwsgi_config behind
is_service_enabled checks.
Change-Id: I997685da771736dbad79bcfe4b00dbc63bd6d6b6
Additionally make the repo name lowercase to match the project name in
our zuul config so that jobs can check the repo out.
Change-Id: Ic2d9c4fa837461bbc29e067a81912b5f72efd3ca
This patch includes changes required to run devstack on CentOS Stream 9
which has been already published in official repos by CentOS team [1]:
- Add RDO deps repository for CS9.
- remove xinetd package from installation for swift. Note that
rsync-daemon is installed which should work fine.
- Replace genisoimage by xorriso in CS9.
- Use /etc/os-release to identify the distro in CS9 as it doesn't
provide lsb_release command.
- Use pip from rpm package instead of from get-pip.py as done in Fedora.
- Add non-voting job devstack-platform-centos-9-stream to the check
pipeline.
Change-Id: Ic67cddabd5069211dc0611994b8b8360bcd61bef
Jobs with OVN_BUILD_FROMS_SOURCE=True are broken
since [1] as ovn nortd not starting due to permission
issues. Fix it by not using sudo for creating OVN_DATADIR
when building from source.
[1] https://review.opendev.org/c/openstack/devstack/+/806858
Closes-Bug: #1952393
Change-Id: I00f0c8c8173b4d8270fbb3e6079d0d8b332e9de5
Before, we needed to unset a couple of parameters that would make the
client return a project-scoped token instead of a system-scoped token,
which we need when interacting with registered limits in keystone.
This commit removes those unsets since we no longer source those
variables by default. This commit also cleans up some of the redundant
parameters in the registered limit calls, like region.
Change-Id: I1af8a168a29e895d57504d41e30efea271ea232d
When initializing OVN, clean up the correct database directory when
using OVN from packages (/var/lib/ovn/ instead of /opt/stack/data/ovn/).
The /opt/stack/data/ovn location is used only when building OVN from
sources, so a fresh devstack deployment with OVN packages may already
have hundreds of existing routers and ports, creating ARP collisions.
Closes-Bug: #1942201
Change-Id: Ic90d4f2f9d8aaef825ea3325c0ad8fef2a1c5e39
The admin_endpoint parameter has been removed from keystone[1], and
setting the parameter is no longer effective.
[1] 192cde56e57a06750641b319da8a72cdcaa554d0
Change-Id: I6ae6a3122668551acc018972624e914fcbb79a22
This user already has the admin role assignment on a project, which
implies the member role, making explicit calls to add the member role
redundant.
Change-Id: I398c5e2f098aeeb896de83872085cbce529a778a
OS_CLOUD is used to communiate to CLI tools what cloud
credentials to utilize.
The change I86ffa9cd52454f1c1c72d29b3a0e0caa3e44b829
unfortunately set an explicit OS_CLOUD account which breaks
any jobs which are expecting a previosuly set OS_CLOUD which
may be different to work. For example, OS_CLOUD being set
as devstack-system-admin to facilitate Secure RBAC testing.
Change-Id: Iee900e552584dda622f57eea3508df48dff2e071
Previously those functions were defined in the neutron's devstack plugin
but with [1] we moved qos related code into devstack and we missed about
moving them too.
This is follow up patch to fix that issue.
[1] https://review.opendev.org/c/openstack/devstack/+/815686
Change-Id: Icf459a2f8c6ae3c3cb29b16ba0b92766af41af30
We can use the devstack-admin cloud configuration everywhere now
and don't need to set environment variables with cloud credentials
any longer.
Fix the swift setup, where some more options need to be explicitly
specified now and the default OS_CLOUD setting overridden.
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I86ffa9cd52454f1c1c72d29b3a0e0caa3e44b829
Those functions were part of the neutron devstack plugin but we
discussed it during last PTG [1] and decided to move to the Devstack
repo plugins which are used by e.g. CI jobs which are defined outside
of the neutron repository.
QoS service is used e.g. in the tempest-slow job which is
defined in tempest and used by many different OpenStack projects.
[1] https://etherpad.opendev.org/p/neutron-yoga-ptg#L142
Change-Id: I48f65d530db53fe2c94cad57a8072e1158d738b0
This commit formalizes some additional users to act as different project
users and updates clouds.yaml file so they're easy to use.
It creates:
- a reader on the demo project
- a reader on the alt_demo project
- a member on the alt_demo project
With the adoption of secure RBAC personas, these are useful for using
OpenStack APIs as that work continues.
Change-Id: I3237a771275311377313b7d7d80ac059ac69d031
IDENTITY_API_VERSION is hardcoded to 3 in most locations already, drop
the remaining occurrences, but keep the variable definition since some
plugins still depend on it. Drop ENABLE_IDENTITY_V2 which no longer
has any effect.
Amend variable list for bootstrap_keystone().
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Change-Id: I06f476d2105bc6ec2b511fc5effcfcc3973eaf97
Keystone has supported system-scope since Queens and we already make
sure we create a cloud profile for system-admin in
/etc/openstack/clouds.yaml.
This commit ensures keystone creates a couple of new users to model
system-member and system-reader personas. Doing this by default in
devstack makes it easier for people to use.
We've already taken a similar approach in tempest by setting up the
various system personas for tempest clients to use.
Change-Id: Iceb7c5f517db20072e121dc7538abaa888423c67
Since Bullseye like Centos 8 Stream needs more memory due to changed
default settings in newer qemu versions, set the swap size to 4G, which
is the same setting already being used for the CS8 jobs successfully.
Change-Id: I83ea34d6aa647d2ab9d4d78ed354904fce836e68
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
If ERROR_ON_CLONE is set to True which is case for
all the devstack based job, devstack does not clone the
repo instead raise error. From current error message, it
is difficult to know that ERROR_ON_CLONE is True until we
traceback the code or check devstack-base job set ERROR_ON_CLONE
to True.
Current error message is like:
-------
+ functions-common:git_clone:560 : echo
'The /opt/stack/oslo.limit project was not found; if this is a gate job, add'
The /opt/stack/oslo.limit project was not found; if this is a gate job, add
+ functions-common:git_clone:561 : echo 'the project to the
$PROJECTS variable in the job definition.'
the project to the $PROJECTS variable in the job definition.
+ functions-common:git_clone:562 : die 562 'Cloning not
allowed in this configuration'
--------
Adding ERROR_ON_CLONE info in error message will help to
know the reason of devstack not cloning the repo.
Change-Id: I9e9852f046fefb299b4ef4446323e9c86437212f
Those functions were part of the neutron devstack plugin but we
discussed it during last PTG [1] and decided to move to the Devstack
repo as plugins which are used by e.g. CI jobs which are defined outside
of the neutron repository.
Placement integration is used e.g. in the tempest-slow job which is
defined in tempest and used by many different OpenStack projects.
[1] https://etherpad.opendev.org/p/neutron-yoga-ptg#L142
Change-Id: Ib86071881f16de1b69c0f9b1b19b6df8b7e66a07
CentOS/RHEL 9 are being compiled for the x86_64-v2 architecture which is
newer than the qemu default of qemu64. This means that for devstack to
boot these instances we need a newer CPU model. Nehalem is apparently
the oldest model that works for x86_64-v2 and is expected to work on
Intel and AMD cpus with kvm or qemu. Switch devstack to this model by
default.
Note that we cannot use host-passthrough or host-model because we want
to support live migration between devstack deployed nova-compute
instances and even within the CI instances that we get the host CPUs can
differ.
Also, we should run this change against as many clouds as possible to
ensure that the newer model works across all of our clouds. There is
some fear that the virtual CPUs presented to us in some clouds may not
be able to run these newer CPU models.
Change-Id: Ibd6e11b59f3c8655bc60ace7383a08458b2177f2
With the depending patch, the endpoint will still be created for
heat tests, so we can turn it off for everyone else.
Depends-On: https://review.opendev.org/c/openstack/openstacksdk/+/777343
Change-Id: I0dc7d6cedd07e942b9f23b26a785b386aff41fbc
The keystone admin endpoint technically isn't different any longer from
the other keystone endpoints in v3 of the API. However, some
applications like heat are still relying on it to exist.
So we make the creation of the admin endpoint during bootstrap optional
here, with the intention to change the default to False once all jobs
that still need this are modified to explicitly require it.
Change-Id: I7ab12141c558186e397c174c248a613d1810011b
Keystone no longer has any special functionality hidden behind the admin
site. KEYSTONE_AUTH_URI which used to point to the admin site has long
ago been changed to be a copy of KEYSTONE_SERVICE_URI, which points to
the public site.
Drop all KEYSTONE_AUTH_* variables except KEYSTONE_AUTH_URI which may
still be in use in some plugins.
This also allows to finally drop the fixup_keystone() function.
Change-Id: I549f3cadc27d137e014241cdd47e90267859c848
The use of this function has been deprecated for a long time[0]. With
PyYAML==6.0 the call is now failing, so replace it with the safe
version.
[0] https://msg.pyyaml.org/load
Signed-off-by: Jens Harbott <frickler@offenerstapel.de>
Change-Id: I7a170262b50a5c80a516095b872d52e1bea5479d
When using glance limits, the create_glance_accounts call needs access
to the devstack-system-admin cloud definition, so we need to create
the clouds.yaml file before that step.
Change-Id: Ie6d807c46b88b16b316aa166870a6a13f2bb346d
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
It may be that it is already compiled in the kernel so there is no
need to load kernel module in such case.
Change-Id: Ie1d32e3fd529e13958857cb3ced6710eebde1e4d
The official Ubuntu cloud images have some further python pkgs
preinstalled that conflict with our requirements. Allow to
overwrite them.
Signed-off-by: Dr. Jens Harbott <harbott@osism.tech>
Closes-Bug: 1871485
Change-Id: I793c250cae5e7b9bc835b7016d790d1f9ae8a7f3
continue is not used in a proper context here (outside of loop). Use
null cmd instead to simply fall through the pip installation.
Signed-off-by: Michal Berger <michallinuxstuff@gmail.com>
Change-Id: Iaea2e5c0177b475edf19d08d71933a74debbb5d9
The current initiator name embedded in our CI images is not unique at
present and can often cause failures during live migrations with
attached volumes. This change ensures the name is unique by running
iscsi-iname again and overwriting the existing name.
We could potentially do this during the image build process itself but
given that devstack systems are not supposed to be multi-purpose this
should be safe to do during the devstack run.
Closes-Bug: #1945983
Change-Id: I9ed26a17858df96c04be9ae52bf2e33e023869a5
Some adaption in database handling is all that is missing. Also add a
platform job that tests this.
Co-Authored-By: Radosław Piliszek <radoslaw.piliszek@gmail.com>
Change-Id: I6dd3e48444dd415d84df5e7f5c74540847cdd6db
oslo.limit isn't currently in the list of libraries that can be
installed from a git repo via LIBS_FROM_GIT.
This adds oslo.limit to enable integrated testing against unmerged
oslo.limit changes.
Change-Id: I26cc567fdf4c84014040ae586bbb029b8de7a236
I83d56c9363d481bb6d5921f5e1f9b024f136044b switched the default of
DEVSTACK_PARALLEL over to True so this dedicated job is no longer
required as *all* jobs should now be using it.
Change-Id: I0f475ab177c2cd49eeb6be861cdd11581e8e0b97
In devstack job, cinder is not a valid service name and logs error
in gate[1] so remove it.
2021-09-28 05:44:47.791807 | controller | + functions-common:service_check:1603 : for service in ${ENABLED_SERVICES//,/ }
2021-09-28 05:44:47.795506 | controller | + functions-common:service_check:1605 : sudo systemctl is-enabled devstack@cinder.service
2021-09-28 05:44:47.809647 | controller | Failed to get unit file state for devstack@cinder.service: No such file or directory
[1] https://e978bdcfc0235dcd9417-6560bc3b6382c1d289b358872777ca09.ssl.cf1.rackcdn.com/801989/7/check/tempest-integrated-storage/779d1e7/job-output.txt
Change-Id: I7ca105201d82b72c7e56778425d3bce7c76047db
Attempting to use LIBS_FROM_GIT="ALL" results in a failure
due to a typo in stackrc for os-resource-classes repo.
Cloning into '/opt/stack/os-resource-classes'...
fatal: protocol ':-https' is not supported
<snip>
[ERROR] /opt/stack/devstack/functions-common:629 git call failed: [git clone :-https://opendev.org/openstack/os-resource-classes.git /opt/stack/os-resource-classes --branch master]
Remove the extraneous '='.
Change-Id: I21f86324dc15fe808b38e366f7af18c96fd3890c
It was required to build MySQL-python bindings but, for some time,
we test and rely solely on PyMySQL which is pure Python and hence
does not require this dep.
This package is going away as distros move towards MariaDB.
Change-Id: I6004ccf28a23009a0fc07bfc9458b59a927b969a
Neutron L3 module in Devstack has way to conigure access to physical
network on the node. It can put physical interface to the physical
bridge or, in case when such physical device isn't set, it creates
NAT rule in iptables.
There was missing the same operation for ML2/OVN backend as L3 agent is
not used there at all.
This patch adds the same to be done in both L3 agent and ovn_agent
modules.
Closes-Bug: #1939627
Change-Id: I9e558d1d5d3edbce9e7a025ba3c11267f1579820
This change fixes the empty value set to the gid option in rsyncd.conf,
which was caused by reference to the invalid USER_GROUP variable, and
ensures the option is set to the group which STACK_USER belongs to.
This also fixes duplicate declaration of the local user_group variable.
Closes-Bug: #1940742
Change-Id: Ifd0a5ef0bc5f3647f43b169df1f7176393971853
This change enables [workarounds]libvirt_disable_apic when devstack is
deployed using the libvirt virt driver and qemu virt type in an effort
to avoid issues outlined in bug #1939108 caused by the older kernel
currently used in Cirros 0.5.2.
Depends-On: https://review.opendev.org/c/openstack/nova/+/766043
Closes-Bug: #1939108
Change-Id: Ibb6c34133bb1c95ef11cc59d9b12a0f65502c61b
... when installed from distribution.
This is mostly to fix Ironic's gate as their ecosystem is too
broad and complex to quickly remove libvirt-python from all
possible requirements.txt
More details inline.
See also: https://review.opendev.org/c/openstack/devstack/+/798514
aka f0bf2bdff1
Change-Id: Ic44daf15e952bbe3c424984ffb2474261e68008f
ironic-tempest-ipa-wholedisk-bios-agent_ipmitool-tinyipa
job is voting on Ironic and neutron gate which mean it is
stable enough and make sense to make it voting on devstack gate too.
ironic-tempest-ipa-wholedisk-bios-agent_ipmitool-tinyipa is alias
job of ironic-tempest-bios-ipmi-direct-tinyipa so using the original
job instead of alias
- https://opendev.org/openstack/ironic/src/branch/master/zuul.d/ironic-jobs.yaml#L784
Change-Id: I95c67ad69e6eae6a72d25a851a71b7de85e56fd2
swift-dsvm-functional job test swift under python3 and
voting on swift gate whihc means this is a stable job now,
let's make this voting to devstack gate too.
Removing swift-dsvm-functional-py3 job as it does not exist anymore
after- https://review.opendev.org/c/openstack/swift/+/731318
swift-dsvm-functional itself is py3 job now.
Change-Id: I58847f74306194eaad132680815101a134fb4022
The MDB backend is the default in Ubuntu and specifying
HDB in debconf doesn't change it to HDB.
Closes-Bug: #1939700
Change-Id: If98f7fc8395678365fb73f0c5cd926cef083e470
Tempest is failing randomly with different reasons
as mentioned in the bug, updating swap size those
issues are not seen.
Before [1] default swap size used to be 8GB but was dropped
to 1G so need to configure it in required job itself.
Did couple of tests in [2] and with 4GB+ swap jobs are
running green. On investigation found that with qemu-5
both Ubuntu and CentOS jobs have memory crunch, currently
Ubuntu jobs are not impacted as they are running with
qemu-4.
[1] https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/750941
[2] https://review.opendev.org/c/openstack/devstack/+/803144
Closes-Bug: #1938914
Change-Id: I57910b5fde5ddf2bd37d93e06c1aff77c6e231e9
keystone has system scope feature implemented since
queens release. Now Devstack also started moving towards the new RBAC.
This commit adds a new job 'devstack-enforce-scope' which enable the
scope checks on service side and see if devstack setting are fine or not.
This job will be expanded to enable the scope checks for the other service
also once they start supporting the system scope.
This will help us to test the scope check setting.
Change-Id: Ie9cd9c7e7cd8fdf8c8930e59ae9d297f86eb9a95
The uninstall here has been around since
Ibb4b42119dc2e51577c77bbbbffb110863e5324d. At the time, there might
have been conflicts between packaged and installed pip. We don't need
it today; get-pip.py keeps itself separate enough in /usr/local on all
platforms. Thus we can also remove the suse/centos special-casing.
python3-pip is in the RPM list so we don't need to
re-install for Fedora.
Add a note on why we are over-installing pip.
Remove some old setuptools workarounds that are commented out.
Change-Id: Ie3cb81a8ff71cf4b81e23831c380f83b0381de71
* update the support distro filter
* don't install xinetd which doesn't exist in F34 any more. I think
there is probably a bit more to do with swift ring-server but that
can be a problem for another time.
* remove old F31 workaround
Change-Id:If2f74f146a166b9721540aaf3f1f9fce3030525c
This uses the python3-pip package for Fedora but maintains the status
quo for existing distributions (i.e. for Suse we run get-pip.py but
don't uninstall, and for everything else we uninstall python3-pip and
run get-pip.py to be running the latest pip).
As noted inline, installing get-pip.py over Fedora 34's package no
longer works, and likely won't ever work again. Unlike the LTS
distributions, the Fedora pip should be more up-to-date, so I think
it's best we just avoid any package overwrites.
Change-Id: I84129aadfcf585bb150a3daa39616246d3d84bbd
Earlier glance remote worker was using same cache directory used by
glance worker. Ideally both should use their own cache directory.
This patch makes provision for the same by setting different path
for image_cache_dir config option.
Change-Id: If2627e9c212fd765b96d925046c04e9cb1001c3d
This reverts commit 9dc2b88eb4.
Reason for revert: Devstack creation/setup the things are not yet moved to scope tokens so we need to wait for that first and then do the scope check enable globally.
Change-Id: If0368aca39c1325bf90abd23831118b89e746222
On some platforms, "python -m pip" isn't available. Currently this is
run undconditionally from the "get_versions" function; remove the call.
Change-Id: I91d6c66d055f02fa7b4368593b629933f82d8117
This reverts commit 7a3a7ce876 and
bcd0acf6c0 and part of
f1ed7c77c5 which all cap our pip
installs.
Given the pip ecosystem can often incorporate major changes, tracking
upstream at least generally gives us one problem at a time to solve
rather than trying to handle version jumps when LTS distros update.
The new dependency resolver included some changes that disallow
setting URL's like "file:///path/to/project#egg=project" in
constraints. Apparently the fact it used to work was an accident of
the requires/constraints mechanism; it does make some sense as the URL
doesn't really have a version-number that the resolver can put in an
ordering graph.
The _setup_package_with_constraints_edit function comment highlights
what this is trying to do
# Updates the constraints from REQUIREMENTS_DIR to reflect the
# future installed state of this package. This ensures when we
# install this package we get the from source version.
In other words; if constraints has "foo==1.2.3" and Zuul has checked
out "foo" for testing, we have to make sure pip doesn't choose version
1.2.3 from pypi.
It seems like removing the entry from upper-requirements.txt is the
important part; adding the URL path to the on-disk version was just
something that seemed to work at the time, but isn't really necessary.
We will install the package in question which will be the latest
version (from Zuul checkout) and without the package in
upper-requirements.txt nothing will try and downgrade it.
Therefore the solution proposed here is to remove the adding of the
URL parts.
This allows us to uncap pip and restore testing with the new
dependency resolver.
Closes-Bug: #1906322
Change-Id: Ib9ba52147199a9d6d0293182d5db50c4a567d677
Now, if no arguments are passed to make_cert.sh script, it will fail on:
tools/make_cert.sh: line 30: [: missing `]'
and might go on with generating certs depending on the bash settings.
It is fixed within this patch.
Change-Id: I62bf9c972ebd1644da622439e05114f245f20809
The default cinder quotas for volumes, backups, or snapshots may
be too low for highly concurrent testing, so make these configurable
in devstack.
Change-Id: Ie3cf3239b48f9905f5760ad0166eea954ecf5eed
rdo-release.el8.rpm rpm points to latest RDO release,
so use it for master, for stable releases use corresponding
release rpm.
Change-Id: I508eceb00d7501ffcfac73d7bc2272badb241494
The usage of sudo with su is not recommended. It results in incosnistent
environment variables. Instead use just sudo with appropriate arguments.
The argument '-u stack' specifies that the sudo will execute as user 'stack'.
The last argument '-i' will launch an interactive shell.
Closes-Bug: #1938148
Change-Id: I42387660480377cdf9a0b04f190e7e1f21fb354f
Starting Wallaby release, nova sanitizes instance hostnames having
freeform characters with dashes. It should be tested in Devstack.
Depends-On: https://review.opendev.org/c/openstack/tempest/+/795699
Change-Id: I54794e58b67620c36e8f2966ec3b62dd24da745b
The glance image size limitation was added and unfortuantely
does prevent larger images from being uploaded to glance. In the
case of all baremetal testing, this value is realistically smaller
than stock "cloud" images which support booting to baremetal with
often requisite firmware blobs, which forces some images over 1GB
in size.
Adds GLANCE_LIMIT_IMAGE_SIZE_TOTAL which allows users who need
larger images to be able to override the default while still
enabling limits enforcement in their deployment. The default
value is 1000.
Change-Id: Id425aa546f1a5973bae8be9c017782d18f0b4a47
As set out in bug #1933096 these bindings are dynamically built against
the version of libvirt present in the environment at build time.
As a result using a pre-built wheel can cause AttributeError's when the
bindings have previously been built elsewhere against an older version
of libvirt installed on the host. This is currently the case in CentOS 8
stream based CI jobs where we try to use 7.4.0 bindings that appear to
be built against libvirt <= 6.10 leading to bug #1933096.
This change seeks to avoid this by installing the bindings from packages
that will always be built against the correct corresponding version of
libvirt.
Change-Id: I76184c17a776c4e1ecaab9549d9d36c8c07c60fa
Closes-Bug: #1933096
The tempest plugin expects the classic environment variables
to be present for credentials to access the cloud, but this is
wrong in cases where we're trying to setup system
scoped services and need to remove the environment variables
that was being used.
Instead, change the plugin to use the os-cloud entry definitions,
and specifically in this case devstack-admin which makes sense
until we begin to start to make tempest itself scope aware.
We likely will want to change the environment variables from being
registered in devstack at some point and completely shift towards
passing an-os-cloud parameter, but that is outside the scope of
this change as doing so will likely break all plugins.
Change-Id: I8d4ec68f116eea07bc7346f939e134fa2e655eac
In RHEL-based distributions, updating setuptools using pip removes the
files from the python3-setuptools RPM. It breaks some tools such as
semanage (which is used by diskimage-builder) that use the -s flag of
the python interpreter (don't import modules from /usr/local).
This commit reinstalls python3-setuptools to fix those applications.
Change-Id: Ib44857e83f75acf37823fae912960a801c83cf7f
Move the 'Starting Horizon' task after the end of the wait for
create_flavors.
The start_horizon function restarts the httpd server, the openstack
services are unavailable during a short period of time, so the
"openstack flavor create" calls might fail randomly.
Closes-Bug: #1932580
Change-Id: I32ee7457586e3de8ba4dfce3b1a12025f9776542
This change add an os-vif lib that declares two new variables
OS_VIF_OVS_OVSDB_INTERFACE and OS_VIF_OVS_ISOLATE_VIF
The former is introduced to workaround bug #1929446 which cause the nova
and neutron agents to periodically block waiting for ovs to respond.
OS_VIF_OVS_ISOLATE_VIF is added to address bug #1734320 when using
ml2/ovs vif isolation should always be used to prevent cross tenant
traffic during a live migration. This makes devstack more closely
mirror reality by enabling it when ml2/ovs is used and disabling it
otherwise.
Related-Bug: #1734320
Related-Bug: #1929446
Related-Bug: #1912310
Change-Id: I88254c6e22b52585506ee4907c1c03b8d4f2dac7
The Secure RBAC effort has updated Ironic such that it
can support a mode where it is scope enforcing for all
interactions with the API. Due to the design, and operating
nature of Ironic's API, services speaking with it must
authenticate with a system scope to have a full picture
of the universe.
In this case, we need to update the nova configuration
accordingly such that the compute service understands
how to talk to ironic so that it can see the nodes under
management.
Ironic will likely update this again at a later point in
time to enable a "hybrid" mixed-mode as the operating model
and related permissions *should* allow nova to use a project
scoped "owner" account with Ironic, in order to access
and command nodes to deploy. But at this time, we're
focusing on the exclusive operating mode.
Change-Id: I1946725ce08c495178c419eaf38829f921c91bbe
Needed-By: https://review.opendev.org/c/openstack/ironic/+/778957
OpenStackSDK has a feature branch "feature/r1" as a preparation for the
R1.0 release. Due to different branch naming functional tests with
devstack are not running. Add temporarily (for the duration of the
feature branch) pragma to the zuul.yaml to allow Zuul run functional
tests. It will be dropped once SDK gets next major release.
Previous attemp didn't work well for other projects, therefore
explicitly include master as well.
Change-Id: I3a5722873f395bc52cc55a0fd6bcea0ebe3b74fc
This table is no longer present on most installations, drop it
from the list to avoid error messages during log collection
that people mistake to be the real error why devstack is failing.
This may lose some debugging information in edge cases, but I
think the improvement of the general user experience is more
important.
Change-Id: Ibb9b247a018a788c8c4b40487762319fe470bf0f
Closes-Bug: 1885198
This fixes various reported and unreported issues with the new
behaviour.
Removes code repetition as well to pay off some technical debt.
Closes-Bug: #1930360
Change-Id: I726c532e96ca434520838ae8a35d5b88b6dd337b
from Tempest to DevStack as it tests DevStack side of things and
is useful for projects not using Tempest.
Verbatim copy except for the devstack- prefix and the /devstack/
path.
Change-Id: Ie166730843f874b9c99e37244e460d7ad33b7eeb
CentOS Stream 8 (aka CentOS 8 Stream) is the currently
supported runtime platform. [0]
Some background history:
The Manila team has asked QA to test centos-8-stream
in the common gate.
A bit later it turned out the point releases of CentOS 8 (aka
CentOS Linux 8) will stop happening entirely by the end of 2021.
[1]
Includes a workaround to the edk2-ovmf issue on CentOS Stream 8
x86_64.
[0] https://governance.openstack.org/tc/reference/runtimes/xena.html
[1] https://lists.centos.org/pipermail/centos-devel/2020-December/075451.html
Change-Id: Iee5a262af757f27f79ba1d6f790e949427dca190
This reverts commit 6843bc798c.
Reason for revert: not sure why but this end up disabling the integration job on check pipeline,
Change-Id: Icfaf8ea17b3ce2e405414c23f8075b18d297bf8b
example: latest recheck on PS12 check pipeline job for neutron - https://review.opendev.org/c/openstack/neutron/+/790060
OpenStackSDK has a feature branch "feature/r1" as a preparation for the
R1.0 release. Due to different branch naming functional tests with
devstack are not running. Add temporarily (for the duration of the
feature branch) pragma to the zuul.yaml to allow Zuul run functional
tests. It will be dropped once SDK gets next major release.
Change-Id: I671b589150fe731125e16316a994a5942219920b
Enable IPv6 private subnet routing in ML2/OVN, it uses the behavior that
already exists in ML2/OVS: add a route from the devstack node to the
CIDRs of the default IPv6 subnet pool. Any IPv6 subnet created using the
default subnet pool and plugged into the default router is reachable
from the host (ex: ipv6-private-subnet).
Change-Id: I02ca1d94e9f4d5ad4a06182f5ac9a2434941cf08
Setuptools v54.1.0 introduces a warning that the use of dash-separated
options in 'setup.cfg' will not be supported in a future version [1].
Get ahead of the issue by replacing the dashes with underscores. Without
this, we see 'UserWarning' messages like the following on new enough
versions of setuptools:
UserWarning: Usage of dash-separated 'description-file' will not be
supported in future versions. Please use the underscore name
'description_file' instead
[1] https://github.com/pypa/setuptools/commit/a2e9ae4cb
Change-Id: I6b8e791c06319fa5fa0935337520c36800b1abd6
This patch moves the OVS compilation module from Neutron into DevStack.
It also renamed it to "ovs_source" to highlight its function, and the
include has been moved to where the rest of the includes are located.
Although this module is not required since by default DevStack installs
OVS/OVN from the host OS packages instead of compiling from source,
this is a nice to have as it avoids having bits and pieces of the code
scattered around multiple repositories.
Co-Authored-By: Lucas Alvares Gomes <lucasagomes@gmail.com>
Change-Id: I39ec9ce0a91bea05cf8c446a9767ab879ac8e8f3
This patch makes the OVN_L3_CREATE_PUBLIC_NETWORK configuration True by
default. This option makes the OVN lib in DevStack create & configure
the external bridge, matching the same behavior from the OVS driver
in DevStack.
Change-Id: Icda53b95fdc3c169ac48a6ec4343c87ba404baa4
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
devstack unit test job does not set any nodeset
and so does use default nodeset defined in base jobs
in opendev. When opendev switches the default nodeset to
the latest distro version, devstack unit test job can start
failing. Example:
- https://review.opendev.org/q/I01408f2f2959b0788fe712ac268a526502226ee9
- https://review.opendev.org/q/Ib1ea47bc7384e1f579cb08c779a32151fccd6845
To avoid such a situation in future, let's set the working nodeset
for this job also so that when we cut the stable branch we can
run it on the working distro version.
Change-Id: I302140778fedf08bc8ba72c453837fa7b8f8f9ae
When Glance is configured with a cinder glance_store, Cinder can be
configured to allow cloning of image data directly in the backend
instead of transferring data through Glance. Expose these
configuration options in devstack to facilitate testing this feature.
Adds:
- CINDER_ALLOWED_DIRECT_URL_SCHEMES
- GLANCE_SHOW_DIRECT_URL
- GLANCE_SHOW_MULTIPLE_LOCATIONS
Change-Id: Iee619b443088fd77cf7b1a48563203bdf4a93a39
The dstat project is no longer maintained.
The pcp-dstat package installs a dstat command so no further updates to
the scripts should be needed.
Change-Id: Ied8c9d29bed4f887c364db7080a0f2a0c02328af
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
This reverts commit 5c304d8176.
Reason for revert: There are more things to fix/move like done in 791085 and 791282 Also let's change all required default in devstack scripts instead of devstack's zuul job side. Basically do this change without any change in .zuul.yaml
Change-Id: Ie0f59d1b9a4b97ad9fd8131819054dfb616f31fd
Sphinx 4.0.0 added a new dependency [0] which is causing the job to fail
at the moment.
This patch fix the problem by adding UC to the docs jobs.
[0] https://www.sphinx-doc.org/en/master/changes.html (LaTeX: add
tex-gyre font dependency)
Change-Id: I28019331017405c06577ada88f8e9f6d9a2afc23
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
As part of the Victoria PTG the Neutron community [0] agreed on changing
the default backend driver from ML2/OVS to ML2/OVN in DevStack. A lot of
changes have been submitted towards this goal including but not limted
to:
* Moving the OVN module to DevStack:
https://review.opendev.org/c/openstack/devstack/+/734621
* Updating the OVN module to use distro packages instead of compiling
OVN from source: https://review.opendev.org/c/openstack/devstack/+/763402o
And now this patch is changing the the actual Q_AGENT,
Q_ML2_TENANT_NETWORK_TYPE and Q_ML2_PLUGIN_MECHANISM_DRIVERS values in
devstack to what is expected by OVN as well as updating the Zuul
templates to enable the OVN services.
[0] https://etherpad.opendev.org/p/neutron-victoria-ptg
Change-Id: I92054ce9d2ab7a42746ed5dececef583b0f8a833
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
The configure_neutron_nova function updates nova configs. While that is
still running we separately update nova configs in stack.sh. This can
result in unexpected configs (that don't work). Fix this by waiting for
configure_neutron_nova to complete its work before we do nova config
updates directly in stack.sh.
For specifics we say that:
[neutron]
project_domain_name = Default
was missing from both nova.conf and nova-cpu.conf and instances could
not be created because keystone complained about not finding domain in
project. The strong suspicion here is that on some systems
configure_neutron_nova would write out project_domain_name while the
stack.sh inisets were running resulting in stack.sh overwriting the
project_domain_name content.
One theory is that disabling swift makes this problem more likely as
there is swift work in the middle of the async period. This is supported
by the fact that our job that hits this problem does indeed disable
swift.
Change-Id: I0961d882d555a21233c6b4fbfc077cfe33b88499
Since victoria cycle, we have moved upstream testing to
Ubuntu Focal (20.04) and so does no Bionic distro in
Xena cycle testing runtime[1]. Grenade jobs also started
running on Focal since victoria was released.
Only thing left was legacy jobs which were not migrated to
Ubuntu Focal in Victoria and as per another community-wide
goal[2], all the lgeacy jobs were suppsoed to be migrated
to zuulv3 native jobs in victoria cycle itself. One of the
pending job was in nova (nova-grenade-multinode) which is also
migrated to zuulv3 native now
- https://review.opendev.org/c/openstack/nova/+/778885
If there is any job running on bionic, we strongly recommend
to migrate it to Ubuntu Focal.
[1] https://governance.openstack.org/tc/reference/runtimes/xena.html
[2] https://governance.openstack.org/tc/goals/selected/victoria/native-zuulv3-jobs.html
Change-Id: I39e38e4a6c2e52dd3822c9fdea354258359a9f53
If a pid disappears on us while we're reading, we should just continue
on.
EnvironmentError is just an alias for OSError since Python 3.3, so use
the latter name. [0]
[0] https://docs.python.org/3/library/exceptions.html#OSError
Change-Id: I3a25cca328e1469f72c84a118a9691c1c0258bc4
Closes-Bug: #1926434
If5c860d1e69aaef9a9236303c370479a7714ad43 attempted to move this default
to lioadm while pinning certain Bionic based jobs to tgtadm.
Unfortunately it missed the legacy dsvm based jobs within various
projects that do not inherit from the devstack-platform-bionic base job
and that are also not covered by devstack's gate.
This change simply forces CINDER_ISCSI_HELPER to tgtadm on Bionic based
hosts to ensure it is always used.
Closes-Bug: #1926411
Change-Id: Ib4b38b45f25575c92fb09b8e97fa1b24af0cc06a
As outlined in bug #1917750 the use of tgtadm in multinode environments
with multiple c-vol services can cause volumes to use duplicate WWNs.
This has been shown to cause some encrypted volume test failures as
os-brick returns a /dev/disk/by-id path to n-cpu that can point to the
wrong underlying volume when multiple volumes with the same WWN are
connected to a host.
There is also some speculation that the duplicate WWNs are also causing
libvirt to fail to detach volumes from instances but as yet this has not
been proven.
This change aims to avoid all of the above by switching the default of
CINDER_ISCSI_HELPER to lioadm for all deployments instead of just EL and
SUSE based deployments.
The Bionic platform job however is pinned to tgtadm as there issues
installing python3-rtslib-fb.
Closes-Bug: #1917750
Change-Id: If5c860d1e69aaef9a9236303c370479a7714ad43
This dumps some data in the case where we fail to wait for a child
pid to help debug what is going on. This also cleans up a few review
comments from the actual fix.
Change-Id: I7b58ce0cf2b41bdffa448973edb4c992fe5f730c
Related-Bug: #1923728
Apparently bash (via POSIX) only guarantees a small (32ish) number of
children can be started and their statuses retrieved at any given
point. On larger jobs with lots of plugins and additional work, we
may go over that limit, especially for long-lived children, such
as the install_tempest task.
This works around that issue by creating a fifo for each child at
spawn time. When the child is complete, it will block on a read
against that fifo (and thus not exit). When the parent goes to wait
on the child, it first writes to that fifo, unblocking the child so
that it can exit near the time we go to wait.
Closes-Bug: #1923728
Change-Id: Id755bdb1e7f1664ec08742d034c174e87a3d2902
Due to issue on stckviz side, job start
failing with POST_FAILURE. If we fix the
issue still we need to wait for periodic job
periodic-package-stackviz-element to publish
the latest tarball on https://tarballs.openstack.org/stackviz/dist/.
Let's not fail the job for any issue occur during
stackviz processing.
Closes-Bug: 1863161
Change-Id: Ifee04f28ecee52e74803f1623aba5cfe5ee5ec90
Ceph adds the osd pool default size option on ceph.conf via [1];
this means we don't need to specify the size of this pool if
the same value (same variable) is used (CEPH_REPLICAS).
This change is an attempt of removing the size setting, relying
on the implicit declaration of the value provided by ceph.conf.
[1] https://github.com/openstack/devstack-plugin-ceph/blob/master/devstack/lib/ceph#L425
Change-Id: I5fa2105ceb3b97a4e38926d76c1e4028f1108d4a
As reported in bug #1920136 the tempest-integrated-compute job has
started to see insufficient free virtual space errors being reported by
c-sch and c-vol when creating volumes. This change simply increases the
default size of the underlying LVM PV used to host these volumes within
the default LVM/iSCSI c-vol backend deployed by devstack.
Change-Id: I965d4a485215ac482403f1e83609452550dfd860
Closes-Bug: #1920136
This patch adds a new environment variable, CINDER_BACKUP_DRIVER for
configuring cinder backup driver used when c-bak service is enabled.
This gets cinder backup driver configurable with a similar pattern to
cinder backends. Although the current configurable backup drivers don't
need cleanup functions, the interface for cleanup is prepared for the
future.
The following backup drivers can be configured:
swift:
This is the default backup driver.
ceph:
This already can be configured if ceph backend driver is enabled. For
backward compatibility, ceph backup driver is used if ceph backend
driver is enabled and no backup driver is specified.
s3_swift:
The s3 backup driver gets configurable with this patch. By specifying
's3_swift', the driver is configured for swift s3api.
In the future, lib/cinder_backups/s3 should be created separatedly for
external S3 compatible storage. This file will just set given parameters
such as a URL and credentials.
Change-Id: I356c224d938e1aa59c8589387a03682b3ec6e23d
Devstack script for setting post-config needs gawk.
So this patch moves gawk from files/*/nova into files/*/general.
Closes-Bug: #1909041
Change-Id: I06a1a5524f146a8d7337963e846b5a6b7561be13
Several jobs have been running in parallel since the late Wallaby
cycle, and other developers have had it enabled locally. I have heard
no async-related stability or debug-ability complaints thus far.
I think that we should convert the default to parallel early in the
Xena cycle in an attempt to spread the speed improvements across the
board, while also collecting data on a wider set of configurations.
Change-Id: I83d56c9363d481bb6d5921f5e1f9b024f136044b
When OVN is built from source, the value of OVS_PREFIX is set to
"/usr/local". All other paths referring to OVS should be prefixed
with this value.
Closes-Bug: #1920634
Related-Bug: #1918656
Change-Id: I9a45a5379d1c47cdf67b9c6d3d0409a88501e61e
In case when OVN_UUID isn't set by user, and it isn't stored
in /etc/openvswith/system-id.conf file, Devstack will reuse it.
If it's not, it will generate and store it in the
/etc/openvswitch/system-id.conf file so it can be set to same value
after openvswitch will be e.g. restarted.
In case when OVN_UUID is set by user, it will be also saved in
/etc/openvswitch/system-id.conf file to make it persistent when e.g
openvswitch will be restarted.
Closes-Bug: #1918656
Change-Id: I8e3b05f3ab83e204bc1ce895baec0e1ba515895b
Instead of doing a flush/add, use replace like the ML2/OVS
code does. Should have the same behavior of not failing if
the address is already present.
Change-Id: If9d8a848b079ccb8c0c9b8e6fb708107aa0d46c7
This cleans up some of the quote and variable handling that was
pointed out in review of the previous patch. This is non-critical,
so I'm putting it in a subsequent patch to avoid disturbing the
careful alignment of patches across three projects that are mostly
approved.
Change-Id: I9b281efd74ba5cd78f97b84e5704b41fd040e481
In order to be able to test glance's distributed import function,
we need to have multiple workers in an arrangement like they
would be if one was on another host (potentially at another site).
This extra worker must be separate from the default image service
in order to repeatedly hit one and then the other to test cross-
service interactions.
This allows you to enable_service g-api-r, which will clone the main
g-api service, modify it to run on a different port, and start it.
The service will be registered in the catalog as image_remote.
Depends-On: https://review.opendev.org/c/openstack/glance/+/769976
Change-Id: I0e2bb5412701d515153c023873addb9d7abdb8a4
Conder started moving to new RBAC and cinder-tempest-plugin
and Tempest need to set few configuration to enable the
scope checks on cinder side and on Temepst side to tell
cinder is all configured with scope checks and test can be run with
scoped token.
Change-Id: Ic7cd919c000c4e7b9a3a06638a5bd87b1617e749
Glance started moving to new RBAC and glance-tempest-plugin
and Tempest need to set few configuration to enable the
scope checks on glance side and so does on Temepst side to tell
glance is ready with scope checks so that test can be run with
scoped token.
Change-Id: I09f513d08212bc80a3a86a750b29b1c6625d2f89
Keystone-tempest-plugin has implemented the secure RBAC
tests and enabling the enforce_scope via keystone devstack
plugin. Doing those setting in devstack will help to manage
easily and in central place also avoid restarting the api
service.
Change-Id: I30da189474476d3397152a0a15c2e30a62d712ad
This file is mega out-of-date and no longer helpful. Remove it.
Change-Id: Ic7e215c3e48a9c453d19355ad7d683494811d2af
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
The XenAPI driver was removed during the Victoria release [1], while the
libvirt+xen driver has been removed in the Wallaby release [2]. Remove
references to Xen from DevStack since its all a no-op now.
[1] I42b302afbb1cfede7a0f7b16485a596cd70baf17
[2] I73305e82da5d8da548961b801a8e75fb0e8c4cf1
Change-Id: If7055feb88391f496a5e5e4c72008bf0050c5356
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
This was removed this service from nova in Ussuri [1]. There's no need
to keep this around.
[1] I2f7f2379d0cd54e4d0a91008ddb44858cfc5a4cf
Change-Id: Idc95c6467a8c6e0c0ed07a6458425ff0a10ff995
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
os-ken is used by neutron ML2/OVS agent.
We need to install os-ken from source to test os-ken changes
against neutron. We already have tempest-integrated-networking job
in os-ken repo but it turns out it consumes os-ken from PyPI :-(
Change-Id: Ibcff212591e9fed25f1316403627269d81455b09
We require the 'tls-proxy' service to set up certificates for us. Hard
fail if 'NOVA_CONSOLE_PROXY_COMPUTE_TLS' is enabled but the 'tls-proxy'
service is not.
Change-Id: I52fec12b78ecd8f76f835551ccb84dfb1d5b3d8a
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
When installing OVN from packages, the rpm for Fedora / CentOS pre set
some configurations that conflicts with the post configuration done by
DevStack.
This patch fixes this problem by erasing the pre-set configuration from
the packages and leaving it to DevStack to configure OVN for its use
(just like we would do when compiling it from source).
Change-Id: I9c18023c9aa79c0633748a6169f4f283e9d74ef0
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
We use Tempest master for testing the supported stable
branches so using master upper constraints works fine but
when we need to use old Tempest in the below cases then master
upper constraints do not work and devstack will not be
able to install Tempest in vnenv:
- Testing Extended Maintenance branch
- Testing py2.7 jobs until stable/train with in-tree tempest plugins
This commit adds a variable to set the compatible upper constraint
to use for Tempest's old version.
Few of the current failure which can be fixed by this new configurable var:
- networking-generic-switch-tempest-dlm-python2
- https://zuul.opendev.org/t/openstack/build/ebcf3d68d62c4af3a43a222aa9ce5556
- devstack-platform-xenial on stable/steinand stable/train
- https://zuul.opendev.org/t/openstack/build/37ffc1af6f3f4b44b5ca8cbfa27068ac
Change-Id: I5b2217d85e6871ca3f7a3f6f859fdce9a50d3946
CentOS 8.3 changed the name of the PowerTools repository to powertools:
https://wiki.centos.org/Manuals/ReleaseNotes/CentOS8.2011#Yum_repo_file_and_repoid_changes
With this repository disabled, DevStack fails to install libyaml-devel,
which causes a failure to install many packages. In my environment
DevStack stopped with an error caused by a missing wget.
Keep the command using the old repository name, for compatibility with
older CentOS releases.
Change-Id: I5541a8aee8467abf10ce8a10d770618bdd693f02
When I reordered the nova database creation for better performance
and cleaner arrangement, I broke the non-standard arrangement where
the super and cell conductors are squashed together. In devstack,
this is implemented by pointing the controllers at cell1 in the
config, which makes it hard to create and sync the databases in the
natural order. This manifested in a failure when running in this
mode (which apparently Trove is).
As a quick fix, this special-cases the setup for cell0 if that mode
is enabled. I will follow this up with a cleaner refactor of all that
stuff so this hack isn't required, but that will take a bit longer.
Change-Id: I5385157c281beb041bf67cba546be20cf9497cbe
Introduced in devstack by I2d7348c4a72af96c0ed2ef6c0ab75d16e9aec8fc and
long tested by nova-next this enabled by most deployment tools by
default now and should be enabled by default in devstack.
Change-Id: Ia76b96fe87d99560db947a59cd0660aab9b05335
This change introduces a basic role to copy the contents of /etc/ceph
between the controller and subnodes during orchestrate-devstack allowing
a multinode ceph job to be introduced by
I9ffdff44a3ad42ebdf26ab72e24dfe3b12b1ef8b.
Note that this role is only used when devstack-plugin-ceph is enabled.
Change-Id: I324c0f35db34f8540ca164bf8c6e3dea67c5b1b4
We have a *ton* of stuff in devstack that is very linear, specifically
the ten-ish minutes we spend loading osc to run a single API command
against something. We also generate configs, sync databases, and other
things that use one core of our worker and make our runtime longer
than it really needs to be.
The idea in this patch is to make it super simple to run some things
in the background and then wait for them to finish before proceeding
to something that will require them to be done. This avoids the
interleaving you would expect by redirecting the async tasks to a log
file, and then cat'ing that log file synchronously during the wait
operation. The per-task log file remains so it's easier to examine
it in isolation.
Multiple people have reported between 22-30% improvement in the
time it takes to stack with this. More can be done, but what is here
already makes a significant difference.
Change-Id: I270a910b531641b023c13f75dfedca057a1f1031
This patchset adds configuration support for network logging
when the OVN driver is enabled.
Depends-On: https://review.opendev.org/768129
Change-Id: I6fc0973bedfd1dcc72b01981cd64f9283662d37c
Signed-off-by: Flavio Fernandes <flaviof@redhat.com>
If we are backing glance with cinder, we will use more volumes and
if timing is right, we will clash with other tests and be unable
to create what we need. If we are backing glance with cinder, we
should increase the volumes quota, which this patch does (to 50 from
a default of 10).
Closes-Bug: #1914665
Change-Id: I2ad1c4d21f996ee1a9ce29ba4f1a4b8f5720f8fb
This commit fixes use of die function before it's defined.
die function can be used after sourcing $TOP_DIR/functions chain-
sources $TOP_DIR/functions-common.
Because fixed portion of stack.sh checks existence of $TOP_DIR/inc
and sourcing $TOP_DIR/function chain source $TOP_DIR/inc, this commit
uses echo & exit command instead of die function.
Closes-Bug: #1913021
Change-Id: I5ec174cf7b02269525b1bfd0bfa94ea889d16fce
On master, we should always enable tempest's verification of Glance's
os_glance namespace enforcement.
Change-Id: Ia71878e6c53ee683a868112959876798e946e2ce
Depends-On: https://review.opendev.org/c/openstack/glance/+/771070
I am still unable to stack because of pip 20.3, but this time
because of the tempest venv build. This forces it to the same
capped pip, which further works around the problem.
Change-Id: Icfaaefe1aa576733764b393cba96d276c9b1cf68
Related-Bug: #1906367
I'm not sure if others will find this useful, but I use this
script to run pieces of devstack while trying to write/debug
things. It saves me a lot of time being able to get to some
project-lib function without a full clean/re-stack.
Figured I'd share in case it's worth putting into the tree.
Change-Id: I9a92fa71d34f50c2f5ba7d11c1a45301bd4478bf
It has been broken for over a month. Feel free to revert in combination
with a fix, better with a commitment to keep the job in working shape
permanently.
Change-Id: I2604374c23716d56de29e16a459b7c7f45b84891
The create_disk() helper had some redundant checks and dead code. This
refactors it to put all the stale cleanup at the top, and groups the
new actions together with more relevant comments to make it easier
to understand.
Change-Id: I1f6218a1994e66786ed9a8065e30bcceec7b8956
Bug #1873234 documents a number of CI failures caused by RPC requests
from c-api to c-vol timing out due to `lvchange` taking longer than the
default rpc_response_timeout of 60 seconds to complete.
While the underlying reason for the slowness should be investigated by
the cinder team a trivial workaround to the fallout created by these
timeouts is to simply double the client RPC timeout used by c-api,
allowing c-vol to return and overall the request to succeed.
Change-Id: I53dc0ae10af6aa13f1349b58373932eb6a15ab02
Related-Bug: #1873234
Since the introduction of I8f24b839bf42e2fb9803dc7df3a30ae20cf264
s-proxy is no longer able to launch as keystonemiddleware (listed under
test-requirements.txt) has not been installed.
keystonemiddleware is listed as extras requirements in swift
- https://github.com/openstack/swift/blob/e0d46d77fa740768f1dd5b989a63be85ff1fec20/setup.cfg#L79
Let's install swift keystone extra requirements also.
Closes-Bug: #1909018
Change-Id: I02c692e95d70017eea03d82d75ae6c5e87bde8b1
This patch changes the OVN module from DevStack to allow for using the
OSapackaged version of OVN instead of compiling it from source.
A new variable called OVN_BUILD_FROM_SOURCE has been introduced and when
set to False (the default value) OVN will then use the packaged version
for setting up DevStack.
Note, in the stop_ovn() function, the OVN metadata agent service name
was wrong and the service wasn't being stopped as part of ./unstack.sh.
This patch also fixed it as well.
Change-Id: Ib41e3b486550200572afd6b3ba783d7644d70d44
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
Co-Authored-By: Slawek Kaplonski <skaplons@redhat.com>
When deploying on the centos 8 stream variant
the output of "lsb_release -i -s" is
CentOSStream instead of CentOS
This breaks the is_fedora function in devstack
preventing package installation and removal.
Change-Id: I39ccefbd06f46adf5077f8d8001f37d3b190f040
This patch caps pip version during bootstrap to avoid the issue:
"ERROR: Links are not allowed as constraints"
A proper fix would be to adapt to new pip behavior.
Depends-On: https://review.opendev.org/764811
Change-Id: I1feed4573820436f91f8f654cc189fa3a21956fd
As documented in the associated TODO this job can be removed now that it
has been migrated to zuulv3 by Ib342e2d3c395830b4667a60de7e492d3b9de2f0a.
Change-Id: Id7c0fd8eec09386cd638956a46c1f75844194b9d
This change updates bionic installs to use the
ussuri cloud archive to enable the use of libvirt 6.0.0.
This is required to prevent a libvirt bug that causes intermittent
failures for the tempest test_live_block_migration_paused testcase.
Change-Id: I9c395c2b5fdfe6ad9a43477280e88e9a9b34f057
Related-Bug: 1901739
At this moment, only image_ssh_user is present in the config
of Tempest. It's set to cirros by default and used for
SSH connections in tests. However, several tests build
instances with image_ref_alt, but still use image_ssh_user to
connect, which results in failure if image_ref_alt is set to
a non-cirros image. They should use image_alt_ssh_user instead,
which can be set to whichever user the image_ref_alt needs in
either local.conf or during plugin installation.
Change-Id: I899909fb71a9862c891e94ba54c6a8fa137f9769
Partial-Bug: #1844535
Right now a system configured with the ceph plugin will not survive
a reboot because the backing disk we create and mount isn't mounted
at startup, preventing ceph from starting and the rest of nova/glance
from working.
This makes create_disk() idempotently write an fstab rule for the
disk we make, and adds a destroy_disk() handler for cleanup.
Change-Id: I50cd4234f51a335af25be756bd2459dca5aa343c
For change 739139 [1] PS 12, the
neutron-tempest-plugin-scenario-linuxbridge died in devstack with
"/opt/stack/devstack/functions-common:237 Failure retrieving default
route device", which comes from
"/opt/stack/devstack/lib/neutron-legacy:237:die_if_not_set".
Looking at the worlddump.txt for that job [2] I see that there is a
default ipv6 route; the vm was not configured with ipv4 networking.
ip route
--------
ip -6 route
-----------
::1 dev lo proto kernel metric 256 pref medium
2607:ff68:100:54::/64 dev ens3 proto kernel metric 256 expires 86380sec pref medium
fe80::/64 dev ens3 proto kernel metric 256 pref medium
default via fe80::f816:3eff:fe77:b05c dev ens3 proto ra metric 1024 expires 280sec hoplimit 64 pref medium
Looking at the devstack code that throws the error [3] it looks like
it only looks for a default route in the output of `ip route`, which
does not include ipv6 information. This change should look in both
the ipv4 and ipv6 route table. A similar check in the L3 setup code
is also updated.
[1] https://review.opendev.org/#/c/739139/
[2] https://d4eb7e3efe98cba79a4b-f4d168cdb20f40841821e4b213645c0f.ssl.cf2.rackcdn.com/739139/12/gate/neutron-tempest-plugin-scenario-linuxbridge/9a6b4f7/controller/logs/worlddump-latest.txt
[3] https://opendev.org/openstack/devstack/src/branch/master/lib/neutron-legacy#L236
Closes-Bug: #1902002
Change-Id: I839e8c222368df98fec308cf41248a9dd0a8c187
(MTU - 50) only supports VxLAN over IPv4, decrease it
to support IPv6 as well, which is 20 bytes larger.
Change-Id: I0cf258770f628c1b4fb590bd274b5433fbcc1450
Sometimes instances don't have an IPv4 default route, so only check for
it when we actually need it. In a followup patch we could extend the
code to check for an IPv6 default route instead or in addition.
Related-Bug: 1902002
Change-Id: Ie6cd241721f6b1f8e030960921a696939b2dab10
This function has been deprecated for a long time, let's finally
remove it. It is only generating a warning anyway.
Change-Id: I7bd440adf2ce8283e3ad3d5d09e6b2b877e2b42e
This patch set the os-vif "ovsdb_connection" configuration option so it
can connect to the local OVSDB. By default, this option points to
tcp:127.0.0.1:6640 and would fail if SERVICE_IP_VERSION == 6.
Also, if SERVICE_IP_VERSION is an IPv6 address, it should be wraped with
square brackets for it to work.
Change-Id: Ie6eec4e140c7464936cf0b0c6307026a94c9f4ee
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
this change corrects the flag order from 'route -6'
to '-6 route' as the -6 flag is an option when used with
ip is an argument to the the ip command and not the route
subcommand.
-6 is accpeted as an argument to the standalone 'route'
commannd but not 'ip route' subcommand.
Change-Id: Ic2ae472e42b7b455693d0aade48dc5109e1f21ba
The "ip route" command only outputs IPv4 routes, add a command to also
show IPv6 route information.
Drop output from "ip link" as that information is contained within the
"ip addr" output already.
Change-Id: Iae87f43c4b1c57f07de041e823da9d350c670389
The current code always overrides tox_environment when running
functional tests, but the correct behavior is to add
the discovered environment variables to tox_environments,
while keeping the user-specified value for it.
The current behavior breaks the devstack-tox-functional children
jobs, like openstacksdk-functional-devstack-ironic, which set
tox_environment.
Change-Id: I5dc9054a1495ca0ef7745c08316441ab153956f4
The situation around glance under WSGI has changed a lot in a week.
We can now run tasks and imports under WSGI, so let's switch the
default back so that glance is consistent (by default) with the
other projects.
Change-Id: I3ae285b2ac4972c0b8abaccfc7c0ede0e1c49bf1
Nova has newer qemu requirements which needs opensuse 15.2 to meet. The
infra opensuse 15 images are 15.2 now to accomodate that. Update
opensuse's supported version to match.
Change-Id: I6f3c5234920b185b2b0cd9c358371402f7a7b922
This includes a workaround to a known dnsmasq >= 2.81 issue that
results in unanswered DHCP requests from instances as documented in the
following Neutron bug:
dnsmasq >= 2.81 not responding to DHCP requests with current q-dhcp configs
https://bugs.launchpad.net/neutron/+bug/1896945
For the time being we will attempt to downgrade to 2.80 to avoid this.
Related-Bug: #1896945
Change-Id: I3a760c43956221424926bd9dad0ebe9b28ae2b52
Enable the compute feature for shelve_migrate on all but LXC and
Xen virt_types.
Related-Bug: #1732428
Depends-On: https://review.opendev.org/#/c/696084/
Change-Id: I31cd00c9117607682213cfa0399709e560f4ad0d
This patch will enable user to configure single cinder store as well as
multiple cinder stores for glance. Below are the parameters needs to be
added in local.conf.
A. For single store
USE_CINDER_FOR_GLANCE=True
B. For Multiple stores
USE_CINDER_FOR_GLANCE=True
GLANCE_ENABLE_MULTIPLE_STORES=True
CINDER_ENABLED_BACKENDS=${CINDER_ENABLED_BACKENDS:-lvm:lvmdriver-1,lvm:lvmdriver-2,nfs:nfsdriver-1,ceph:cephdriver-1}
GLANCE_CINDER_DEFAULT_BACKEND=lvmdriver-1
enable_plugin devstack-plugin-nfs https://opendev.org/openstack/devstack-plugin-nfs
enable_plugin devstack-plugin-ceph https://opendev.org/openstack/devstack-plugin-ceph
NOTE:
GLANCE_CINDER_DEFAULT_BACKEND should be one of the value from CINDER_ENABLED_BACKENDS.
If you need to configure nfs and ceph backend for cinder then you need to add respective plugins in
local.conf file.
If GLANCE_ENABLE_MULTIPLE_STORES is True then it will not configure
swift store for glance even if it is enabled in local.conf file.
Needed-by: https://review.opendev.org/#/c/750018
Change-Id: Id0d63c4ea41cce389eee8dc9a96913a7d427f186
nova-ceph-multistore is defined in nova side to test the
glance multistore on ceph and it is voting on nova gate.
There are other multistore testing enhancement going on for
example- https://review.opendev.org/#/c/743800/. so to avoid
any regression, let's run exiting mutistore job on devstack
gate too.
Change-Id: Ie82b4057463df4b6138c53b14a582bd84866aebd
There is flag Q_BUILD_OVS_FROM_GIT which can be used to not compile
ovs from source.
But this wasn't respected in the ovn_agent's module in install_ovn
function which was always installing from source ovn and ovs.
We need to disable that e.g. on grenade jobs when new version is
installed.
Change-Id: I7d3f92365e880191dcfe7c618a6f79d5f741144f
This is will allow the openstack/nova project to facilitate a minimum
required version bump of QEMU and libvirt within the libvirt virt driver
in I8e349849db0b1a540d295c903f1470917b82fd97 ahead of the planned switch
to focal later in Victoria.
Change-Id: I85eb45632ff229676f7c29708f4a7cc64b3d90e3
Previously pyc files were only cleaned if clean.sh was run.
with this change a new clean_pyc_files function was introduced
with the logic that was previously in clean.sh but it is now
invoked from unstack.sh
With the previous behavior you could not stack with horizon
enabled then unstack and stack again due to the presence of pyc
files that were owned by root.
By moving the clean to unstack in stead of clean.sh you can
now stack, unstack and stack again without hitting the pyc issue.
since unstack is invoked by clean the existing clean.sh behavior has
not changed in practice except for the fact the pyc files are
removed sooner in the process.
This change also removes support for findutils < 4.2.3
Ubuntu 12.04 and CentOS 6 both have 4.4.2 since they were
released 8 years ago and are now EOL its fair to assume
that all modern distros have 4.2.3+
https://repology.org/project/findutils/versions
Change-Id: I13c9aad9be7e0930a0d875b7d382090caf0b8982
This reverts the changes made to the zuul config in [0]. Since the
backports to the stable branches have been merged, the jobs can be
voting again.
[0] I5d8aa0e58e0409c54451b51de5eb70ba9a68d849
Change-Id: I3f5972e05faea8f11d7e87f3f8b05e4979e6c328
A new setuptools release has changed the way pip installs are done,
see [0]. With this change we switch back to using the distro
method for global pip installs.
Temporarily make grenade jobs non-voting in order to allow this
patch to be backported.
[0] http://lists.openstack.org/pipermail/openstack-discuss/2020-August/016905.html
Change-Id: I5d8aa0e58e0409c54451b51de5eb70ba9a68d849
This patch is a follow-up of Ib4194329474e8d68a90886d2a04f027eecd741df.
This patch removes the configure_port_forwarding call from the
neutron-legacy module because port forwarding (just like other
extensions such as DNS, QOS, etc...) are already enabled in the
plugin.sh file in the neutron repository [0]. The
configure_port_forwarding method itself is also defined in the neutron
repository so calling it here may result in a failure in case the plugin
is not enabled.
We are also removing the "dns" extensions from the default
Q_ML2_PLUGIN_EXT_DRIVERS variable because this extension conflicts with
the default DNS extensions that is enabled by Neutron when
q-dns/neutron-dns service is enabled (also in [0]). The LP for this
conflict problem is: https://bugs.launchpad.net/neutron/+bug/1887163.
[0]
https://github.com/openstack/neutron/blob/945a244588b81064e4301b6f055a3c90f472bd7e/devstack/plugin.sh#L101-L103
Change-Id: Iafb9e45520798b2a612192cfc6cca28501465862
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
This patch fixes an early stack issue where the following error message
would be presented:
ERROR: Cannot uninstall 'PyYAML'. It is a distutils installed project
and thus we cannot accurately determine which files belong to it which
would lead to only a partial uninstall.
We also drop references to removal of Python 2 library egg infos now
that Python 2 is EOL.
Closes-Bug: #1892363
Change-Id: I2876ee58ab6b73682869d6b4e684e10ac5e56ad9
The role reads the OS_* variables set by devstack through openrc
for the specified user and project and exports them as
the os_env_vars fact. It is meant to be used as a porting aid
for the jobs of the non-unified *clients only.
This is useful to run functional tests against the generated
environment. A slightly less generic version of this role
is currently used by python-cinderclient and python-novaclient
(get-os-environment).
In order to make this more useful, call it from
devstack-tox-functional and derived jobs. The role
execution is disabled by default and it can be enabled
by setting openrc_enable_export: true.
Change-Id: I15f5a187dbc54c82e8f4a08b4bb58d56e0d66961
There are a few dependencies which are really os-brick-specific.
They are listed in its bindep.txt file, but os-brick is usually
installed from pip, so its bindep.txt file is not available.
As those dependencies are needed by the various services
which use os-brick, move them to their own dependency file
(with the addition of the new lsscsi, required by the next
os-brick stable release) and make sure that file is parsed
when installing the services which require os-brick.
Side note: there should be a way to avoid this duplication
also for pip-installed libraries (normal services can use
files/ or even bindep, but in this case the source is not always
available), (temporarily?) duplicate them, as it has been
the case for the other os-brick dependencies already listed here.
Change-Id: I9ab6e215dbef9ebdb1946da2f9a40ce020ecc95b
As part of the Victoria PTG the Neutron team entertained the idea of
having the OVN driver as the default backend in DevStack (this hasn't
yet being decided by the community, this will be discussed within this
cycle).
For this to happen, we also would need to move the module that configures
OVN to the DevStack repository. This is what this patch is doing.
Note that we are updating the lib/neutron-legacy module instead of
lib/neutron in this patch, this is because as part of the PTG the
Neutron team has decided to un-deprecate the neutron-legacy module since
the "new" lib/neutron module is broken and nobody is current working on
it (also all services uses neutron-legacy).
Also, the ovsdbapp has been added to the ALL_LIBS list because a gate
job in the ovsdbapp project repository relies on installing the library
from source instead of pip to run.
Depends-On: https://review.opendev.org/#/c/740663/
Change-Id: Ib4194329474e8d68a90886d2a04f027eecd741df
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
We need to use python3, our deps are no longer installed in python.
Includes the current set of updates to the plugin registry.
Change-Id: I4753ddf60ed066cc11fa74dbbd63210dbad733a8
While RHEL 7 is detected as RedHatEnterpriseServer, RHEL 8 is
RedHatEnterprise.
$ lsb_release -i -s
RedHatEnterprise
Change-Id: I3d750d808c6ebea9c230f0508cdbc016415b9922
Most libs maintain their own system packages in a local bindep.txt file.
We don't currently use those when installing packages from source, which
can result in broken package installs.
This adds a flag to always attempt to install bindep packages if the
bindep.txt file exists. If a file cannot be found, it will just emit a
warning and carry on.
Change-Id: Ia0570f837b8af1c3fee0a314b026a4a7ed27e6a9
Signed-off-by: Sean McGinnis <sean.mcginnis@gmail.com>
If glance is in standalone mode, image import works
fine so enable the tempest tests. Once we will have image
import or other async tasks working with glance under uwsgi
then we can remove this flag and run import tests by default.
Depends-On: https://review.opendev.org/#/c/741425/
Change-Id: I853e8a3815187f0aa8f05c70488ec948a97e55a6
As of the referenced patch in glance, we can do import in wsgi mode.
Also remove the enforcement that import methods are disabled.
Change-Id: I8da4b4ad6105bb64c4045ca80db9742591d01564
Depends-On: https://review.opendev.org/#/c/742065
We always want to start glance on the internal port now,
regardless of whether or not tls-proxy is in use, because we
write_local_proxy_http_config() for the standalone case.
Change-Id: I47dea645d4a852e02e25af0e1df9c28fec92c42a
Co-Authored-By: Radosław Piliszek <radoslaw.piliszek@gmail.com>
While removing registry [1] we by mistake removed some code related to
multiple store configuration for glance. This must be happened during
resolving merged conflicts.
Adding it back.
[1] https://review.opendev.org/708062
Change-Id: I2b84f7b7c51b7b20765a06b48c75006fd2e8ab71
Glance should not be exposing import methods that cannot work via its
API, but it does today. In order for tempest (et al) to be able to
properly detect whether import is possible, we must configure the
import methods in standalone mode, or disable them in wsgi mode. The
referenced Glance patch will make this a requirement.
Change-Id: I3bf3498d83607c5e98b70877c061dc54fc3c0a6e
Needed-By: https://review.opendev.org/#/c/741497/
A whole set of Glance functionality is not usable under uwsgi, including any
of the more powerful async import, customization, and copying functions.
In order to facilitate writing and running tempest tests for these features
in all environments covered by the various jobs across all the projects that
include Glance, we should default to this deployment method.
It is still possible to deploy glance in uwsgi mode by setting the flag to
False, and we can do that for some jobs to make sure that it continues to
work. However, the default should be what we expect deployers will use,
which is standalone mode.
Depends-On: https://review.opendev.org/741479
Change-Id: I141acab2a07a4eebd8d850f900058bc8cbf9c7bf
This makes no sense to set on _every_ devstack deployment, only
if we are using libvirt qemu or kvm. Make it conditional.
Change-Id: I860e899274646ff73b8f084a0583325239aee9cc
This abstracts out the conversion of key=value,... property lists to a function
and makes both _upload_image() and the two ami/aki image create calls use it. The
move to bare key=value properties introduced a regression for ami/aki where
the --property flag stopped being passed to osc in that case.
Change-Id: Idf7fdfe3f5800f79f6c48f9d9606a7b53436a730
Full Glance functionality requires Glance being run in a configuration
where it can spawn long-running task threads. The default uwsgi mode
does not allow this, and the current workaround is to set WSGI_MODE
to something other than uwsgi to get the devstack code to deploy
Glance as a standalone service. Since this affects the entire rest of
the deployment, this patch separates out a flag to control this behavior
specifically for Glance. When WSGI_MODE=uwsgi, control of the Glance
deployment mechanism is allowed via GLANCE_STANDALONE=True|False. If
WSGI_MODE!= uwsgi then we deploy standalone Glance anyway.
Change-Id: I79068ce0bd7414bc48ff534ee22f0de5d7b091cb
Added new boolean option 'GLANCE_USE_IMPORT_WORKFLOW' default to False.
If this parameter set in local.conf as True then devstack will use the
new import workflow to create the image.
In order to use new import workflow of glance;
user need to set below options in local.conf
GLANCE_USE_IMPORT_WORKFLOW=True
Note that the import workflow does not work in uwsgi because of
some fundamental restrictions it has. Thus, devstack must be configured
with WSGI_MODE=mod_wsgi, otherwise glance will not be able to process
the imports. The new helper function will abort to avoid in that case
to avoid the image never being moved to "active" state by an import
task that will never be executed.
Co-Authored-By: Abhishek Kekane <akekane@redhat.com>
Co-Authored-By: Dan Smith <dansmith@redhat.com>
Needed-By: https://review.opendev.org/#/c/734184
Change-Id: I1306fe816b7a3eca1e2312c0c454be3d81118eca
Replace legacy-tempest-dsvm-lvm-multibackend with its native version,
cinder-tempest-lvm-multibackend.
Remove legacy-tempest-dsvm-neutron-dvr-multinode-full, which was defined
as an experimental job in neutron and removed during the ussuri lifecycle.
See https://review.opendev.org/#/c/693630/
Change-Id: I76d1efaa3a6c1fe9825e8457438f514114b2ecad
The old peakmem_tracker service has been disabled in [0], now enable
the replacement memory_tracker.
Also fail when the old service is still configured, otherwise
consumers might never notice.
Depends-On: https://review.opendev.org/739995
Change-Id: I583caf3f36a8ff41d7d4106dabc6c5f24243085e
Starting up n-api-metadata service on every compute nodes does
not solve the problem of isolated networks (no route to metadata
service). It all depends on how 'enable_isolated_metadata' and
related options (e.g. force_metadata) are set in dhcp agent and
what is configured for the 'nova_metadata_host' option of q-meta
service. Having a global n-api-metadata service in the control
node is sufficient for a mult-node lab setup.
Besides, the n-api-metadata services on compute nodes are not
really working due to https://bugs.launchpad.net/nova/+bug/1815082
Change-Id: Ib8691c3eeee59758fbd98989d9460f1458ea422f
Related-Bug: 1815082
This reverts commit f6286cb586.
This patch is blocking glance as it needs mod_wsgi to perform new import workflow.
Change-Id: I4475247dfe986114d37678b3d3d552c0c7d02ddc
Recent regression spotted by Dmitry Tantsur.
DevStack dropped Py2 support but the now-unused-in-devstack
python3_enable got its result nastily inversed.
Change-Id: I4b37cc847a24705c4955cec2e6e45f0514705f1b
The non-voting flag was accidentally dropped in [0].
[0] Ib4416dc2f5e003fd770f5240a8f78213c56af8e6
Change-Id: If9519f1ac9afd66553e1c1410fdc16369f166b98
No targetcli package on Ubuntu Focal, it should use targetcli-fb also
when "$CINDER_ISCSI_HELPER" == "lioadm".
Although Xenial only has targetcli, but Xenial is dropped from CI. And
starting from bionic, Ubuntu uses targetcli-fb to replace targetcli. So
here we can use is_ubuntu to make ubuntu use targetcli-fb.
Change-Id: I6d35b6651d486e716980dcd9f4d693bed560463a
This option sets enable_v1_api in glance-api.conf, a setting that
was removed by change Ia086230cc8c92f7b7dfd5b001923110d5bc55d4d
in July 2018, so remove the devstack option from lib/glance.
It occurs in two other places:
This option is used in lib/cinder to set an option that was removed
from Cinder by change Ice379db9ae83420bacf9e96e242c7515930eae86 in
Queens, so remove the related code.
When this option is False, it is used in lib/tempest to set
[image-feature-enabled]/api_v1 to False in the tempest config file.
However, the default value of ths setting has been False since
change Iab3a209c744375bf2618afc00a3f7731b62f557e in Sept 2018,
so remove the related code.
Change-Id: I4b18a0a388ed7e7a392fabeac613778e0d23dee7
Those historic references to port 5000 and 35357 aren't being used
anymore for some time, so let us drop them.
Clean up some python2/3 wording along the way.
No longer mention Identity API v2, which is also a thing of the past.
Change-Id: Iafff097eee082f24ea2ae27ad038ad115aa36c61
Keystone no longer has any special functionality hidden behind the admin
endpoint. Stop referencing it in consumers, so it can later be dropped
completely.
Change-Id: I04a5d77908005268cc7c59e7e9ddeea70f6732e2
This commit adds the focal nodeset for
multinode so that those can be used for
projects side jobs or testing.
We need to define these as first step to avoid
any conflict on nodeset name if project started
defining these.
Example: three node focal nodeset is already
defined in x/tobiko, fixing the same in depends on.
-I30a6bb63269f031a74f9bff6c765d59d91088797
Depends-On: https://review.opendev.org/#/c/738128/
Change-Id: I5ce49f7a7d52d00555c14b08864bc8975956b20c
Story: #2007865
Task: #40212
Some arping versions only accept an integer number for the
"deadline" (-w) parameter.
Change-Id: Ie21c9b5820262d049c0fcd8147d85cc110d88272
Closes-Bug: #1885169
Dragonflow was retired in 2018 and is now being retired. This removes a
documentation reference to it.
Change-Id: I24ab79482306a7c816b5242a981f1b508ff8f6ec
Signed-off-by: Sean McGinnis <sean.mcginnis@gmail.com>
We haven't had a working job for stretch or jessie in years, attempts
to get things fixed have been dropped, set let's be honest and drop
those from our support list.
Change-Id: Ia6152be79f8044f7ff039ec0911ad4938d6271f4
- Add a nodeset and a platform job
- Drop uwsgi-py2 pkg that no longer exists
- Blacklist tests that are currently failing
Change-Id: Ib4416dc2f5e003fd770f5240a8f78213c56af8e6
Focal only provides a pip3 binary, no pip3.8. Instead of working around
that with a symlink, use the module instead.
Add version information output for this variant.
Change-Id: I7af194ecc40e4d43c10ce067a661bb6ab4ca37d4
process-stackviz role currently defined in Tempest repo[1] and used
in tempest jobs. Issue with having it in branchless Tempest is that any
fix in that role cannot be backported to older Tempest. Also stackviz is
not Tempest things it is only used by Tempest jobs.
Stackviz can be considered as a service similar to Tempest so inatlling it in
Devstack side make sense. Also that will give us advantage to handle the
branch specific fixes or backpoting the fixes.
This can solve the current issues we are facing on rocky branch
- https://zuul.opendev.org/t/openstack/build/c1e2da80537448bfb24331fb4198a931/console#4/0/29/controller
[1] https://opendev.org/openstack/tempest/src/branch/master/roles/process-stackviz
Change-Id: I0ba1fd05c6391cd8bb978119ecfbb891def2d497
During the Victoria cycle we plan to switch default Neutron backend in
Devstack from OVS to OVN.
As first step before we will start discussion about this change with
whole community, we want to add tempest ovn based neutron job to
the devstack check queue so that we can keep verifying that with
devstack changes.
Change-Id: I8484baa7398b28ed5ef62f86b55022c7d8703f56
Glance has deprecated registry serivce for long and now efforts are placed to
remove the registry code from the glance repo.
To avoid regression on other projects, gate jobs etc. removing
configuring registry service from the devstack.
Change-Id: I6a7be6bdc97acc43c8e985060aeea05d92642e80
Neutron-fwaas is going to be deprecated in master branch with [1].
[1] https://review.opendev.org/#/c/735828/
Change-Id: I513ef36e681fc3f9e5aa9f81c9aedba716366729
Building uwsgi from source was a workaround that was introduced a long
time ago, it doesn't seem like it is needed anymore and will actually
fail for Ubuntu 20.04.
Also it doesn't match what will happen for most real-world
installations, so let's try to get back to using distro packages. We'll
still use the source install for RHEL/Centos, it remains to be tested
whether we can get back to using distro versions there, too.
Change-Id: I82f539bfa533349293dd5a8ce309c9cc0ffb0393
uwsgi broke installation from source with their latest release [0].
Since we want to move away from source installation anyway, make
grenade based jobs non-voting for the moment so that we can backport
[1] properly.
[0] https://bugs.launchpad.net/bugs/1883468
[1] https://review.opendev.org/577955
Related-Bug: 1883468
Change-Id: I8e47bb7c70031a4df7f1af6b811df4c6cc784b2a
Tempest option scenario.img_dir has been deprecated more than 4 years, it's
time to remove it from devstack.
img_file option should contain the full path to the image.
This patch removes setting of img_dir option and makes img_file
one contain the full path of an image.
Change-Id: I71102095f3603915f0bc7d21f2e18c4eac4e95ec
Depends-On: https://review.opendev.org/#/c/710996/
Related-Bug: #1393881
The FLOATING_RANGE variable should specify the network address of the
used prefix for clarity.
Change-Id: I547bd42d8bdc5f0f2001d47f2d5b43729773b1bc
Closes-Bug: 1870204
This is already unconditionally installed via install_apache_uwsgi in
stack.sh; we don't need to install it again in keystone. Since we
need workarounds on some platforms (see
I3bc5260e77cebe852cc8d70d9eddf84ef71d74bb) we only want to do this in
one place.
Change-Id: I40d84cbdf68cf6bb5cba143b6c0c126cdb8a84d4
All these uwsgi invocations assume that the uwsgi binary is in the
same directory as their project binaries are installed into (probably
/usr/bin). That may not be correct -- for example if using a packaged
uwsgi on Fedora the binary will live in /usr/sbin/uwsgi (not /usr/bin
where the project files from pip are).
Switch invocations to just find it in the path.
Change-Id: I298e3374e9c84e209ffcabbaaacda17f8df19f4f
Centos 7 is no longer supported, replace with Centos 8. Also Debian
hasn't been working for some time and progress in fixing it has stalled,
so drop it for now.
Change-Id: Ic1513b20f296978bca095c7aa79f022d7d9ab7ac
Closes-Bug: 1881183
This option was deprecated in Ussuri [1] as nova no longer provided any
schedulers aside from the filter scheduler and the existence of
third-party schedulers was very unlikely. Stop configuring this and
simply use the default.
[1] https://review.opendev.org/#/c/707225/
Change-Id: Iabdd1d00e00ee269334f0fe0db265a97207e2dc6
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
We need an appropriate auth_uri for s3token to be able to contact
Keystone.
Since tempauth is always enabled, we want to delay the auth decision
until after tempauth has had a chance to try.
Change-Id: Ie4ff33a617b9dc74d51d037ec8ebd0d9787dd76d
For a long time, swift3 recommended a pipeline like
... swift3 s3token authtoken keystoneauth ...
This led to inefficiencies where the proxy would first contact Keystone
to validate the S3 signature and issue a token, then contact Keystone
*again* to validate the token ID that was just issued.
After s3token moved into the swift3 repo, it was improved to be able
to put all of the headers into the WSGI environment that Swift's
keystoneauth middleware expected and the recommended pipeline was
changed to something like
... authtoken s3api s3token keystoneauth ...
At the time, the old order would still work, it would just be less
efficient. When support was added for Keystone v3, however, the new
order became mandatory.
All of that happened before swift3 moved back into Swift as s3api, but
the pipeline placement problems are the same: Keystone users won't be
able to use the S3 api with the current order.
Change-Id: Id0659f109cc2fc12ddb371df0b26812ba8c442d9
Related-Change: I21e38884a2aefbb94b76c76deccd815f01db7362
Related-Change: Ic9af387b9192f285f0f486e7171eefb23968007e
The bug has been fixed since a while, also in recent distributions,
for example Ubuntu 20.04, the sgabios.bin ROM is provided directly
by qemu-system-data as an actual file under /usr/share/qemu and
it conflicts with the one provided by sgabios, so removing the
workaround is actually needed to prevent failures.
Change-Id: Ib5f23dbd8839a0927418692054f4ed4abd76babc
Switch to openstackdocstheme 2.2.1 version. Using
this version will allow especially:
* Linking from HTML to PDF document
* Allow parallel building of documents
* Fix some rendering problems
Update Sphinx version as well.
Set openstackdocs_pdf_link to link to PDF file. Note that
the link to the published document only works on docs.openstack.org
where the PDF file is placed in the top-level html directory. The
site-preview places the PDF in a pdf directory.
Set openstackdocs_auto_version to not auto-version the documents.
Set openstackdocs_auto_name to use 'project' as name.
openstackdocstheme renames some variables, so follow the renames
before the next release removes them. A couple of variables are also
not needed anymore, remove them.
Change pygments_style to 'native' since old theme version always used
'native' and the theme now respects the setting and using 'sphinx' can
lead to some strange rendering.
See also
http://lists.openstack.org/pipermail/openstack-discuss/2020-May/014971.html
Change-Id: I26887b175b9c1ced7347289b7d5d4f57a20ec36a
tmpfs exhaustion has long been suspected as the root issue behind
failures to load ssh keys and other metadata from local config drives as
documented in bug #1808010. This can also lead to failures fetching
metadata from n-metadata-api leaving Tempest unable to SSH into
instances.
This change increases the RAM of the m1.nano and m1.micro flavors by
64MB to hopefully avoid these errors going forward. This is also ahead
of our eventual upgrade to Cirros 0.5.0 where 128MB becomes a
requirement.
Related-Bug: #1808010
Change-Id: I4b597579cf89939955d3c110c0bd58ca05de61f0
Disable insecure option for glance_store with
swift backend when tls is enabled.
Specify swift_store_cacert option.
Change-Id: Ia1e8f596c95dd7b6e63cb21a94c8316dc71bf945
Follow the pattern of the other configuration keys. The new variables
allows tests to enable/disable volume revert tests provided
by cinder-tempest-plugin.
Revert-to-snapshot was introduced in pike, and so the tests.
Change-Id: If137f201c2f646703f5a1ff96e71e48caed63b67
Ubuntu Focal doesn't have python-pip, only python3-pip. Trying to
uninstall a package that apt doesn't know about (installed or
uninstalled) results in a nonzero exit code so devstack fails. This
patch makes the package removal safer for both python2 and python3 cases
by checking first if the package exists.
Change-Id: I3b1118888cb0617ffb99b72c7e9a32308033783e
This repo is now testing only with Python 3, so let's make
a few cleanups:
- Remove obsolete section from setup.cfg
- Use newer openstackdocstheme and Sphinx version for python 3
- Remove install_command from tox.ini, the default is fine
- Move basepython into testenv
- Update bashate version
Change-Id: I3d78b3787af2efce831d223dbcab6cf84c358028
The integrated gate template (integrated-gate-py3) has been switched
to the new grenade name (grenade-py3 -> grenade). This repo uses the
template but also has for irrelevant files an extra entry.
Rename the job following the template change to avoid duplicate
grenade runs.
Details:
- http://lists.openstack.org/pipermail/openstack-discuss/2020-May/014602.html
Depends-On: https://review.opendev.org/725148
Change-Id: I73e67c0e01ca231678903b2460dab672f17251e3
We no longer support platforms with Yum on master. Cleanup old
references and convert to dnf.
We don't need any of the failure wrapper stuff as dnf runs in strict
mode by default.
There seem to be a few callers out there, so we'll leave it called
yum_install for now.
Change-Id: Ie71a48fd85b00a97a14bf260cd013b18af4cce06
This package isn't available on some distributions hasn't been
required in Neutron for several years
If679e79fa3242ee1cd8610b5525deca35b41c87e. Remove it.
Change-Id: I7308a885c1d084efe2f0b9f542443d35966140ed
The paths for the devstack source directories are currently hardcoded,
this patch allows overriding that list.
Change-Id: I2b673b0d110d84658b89bb14663584330deaf3aa
With the removal of CentOS 7, we can re-evaluate the rpm
installations.
We should not need virtualenv after https://review.opendev.org/712609
There should be no need for python-devel as we're python3
pyOpenSSL was added to workaround memory issues in
9e98f9435e (2015) ... I think we've
moved on.
pyxattr is not a package; remove it.
I don't see we need packaged m2crypto, which isn't a package on CentOS
8. nor libxml2-python; these days it has wheels which should work
with the normal installation process.
centos8 has:
* targetcli
* pcp-system-tools (and no dstat)
* iptables-services
* java-1.8.0-openjdk-headless
* kernel-modules
* rsync-daemon
just as all supported fedora's do, so we can remove any matching here.
Change-Id: I542c426a67a98f331d2a29bacd220af81fab8cc4
We have lib/databases/mysql which is installing databases, remove it
from the bulk package lists.
Split is_fedora (fedora & centos8 -- soon) to install mariadb-server
and mariadb-devel to retain status-quo.
On suse this seems to be a meta-package
'mariadb-server' not found in package names. Trying capabilities.
so split that out. It seems it has never been installing the -devel
package, and things work (presumably clients are coming from wheels so
don't need to build against it).
Change-Id: I86433318e8f76c40c5c792b795411a5c9d8351d3
We do not support CentOS 7 on master branch due to no Python 3 or
ongoing eco-system (i.e. RDO) support; see
Id9ef507dd6f4226d65c6ed3043666b0aa6a3bd1c.
Change-Id: If98581708568e7a8d15e6edc588a008df0cac0fb
This includes the addition of the python3-virtualenv package required to
provide the virtualenv binary that is no longer present in the image.
Change-Id: Ie8e66d8b9f93063b97f88f41a626daddf235339b
Using venv, which is part of python3, we avoid an extra dependency on
the virtualenv package. For Debuntu, which splits this out into a
separate package, add this to debs/general.
This is part of the infra efforts to ship "plain" nodes without any
dependencies installed. While devstack can re-install virtualenv, we
don't need any features it provides and it means one less dependency.
Change-Id: I3c323640f288e57581a4eb8adba2a08d0b0cbd8f
There's a bug[1] with the combination of the p11-kit and
ca-certificates-mozilla packages available on the latest built
opensuse-15 node in nodepool (which has not been rebuilt for weeks due
to a separate issue[2]) which causes the standard CA bundle to not be
installed correctly and causes jobs that call to external HTTPS services
to fail. Upgrading both packages in sync fixes the issue.
[1] https://bugzilla.suse.com/show_bug.cgi?id=1154871
[2] http://bugzilla.suse.com/show_bug.cgi?id=1166139
Change-Id: Ia8fdfe12fd9089e178adcb2b5eec997eebada262
Needed-by: https://review.opendev.org/713566
When stacking outside of Zuul CI the wheels have to be built locally and
python3-dev package is required.
Story: 2007491
Task: 39213
Change-Id: I0960269d5cf193c9ececc5490485522c74646382
This first came in with Id749c37ab7fefa96b35f11816b56b9def5ef4b08. It
talks about ancient versions of pip; can't see we need it any more.
Change-Id: I9d4831955070990a81a809d988612d9d5b1aa672
Recent change to devstack dropped installing test-requirements [1]
However, this caused gate failures due to lack of glance-store
deps for cinder and swift support.
This patch makes devstack install relevant extras depending on
enabled features.
Additionally, relevant functions are added/fixed to make this
possible.
glance-store = glance_store (for gerrit search match)
[1] https://review.opendev.org/715469
Change-Id: I0bf5792a6058b52936115b515ea8360f6264a7c9
Swift keeps testing py2 but we should keep both in shape.
To fix stestr on py2:
Depends-on: https://review.opendev.org/715942
Change-Id: I616e39c64e22d467d7186dba98226cc5beef23ea
This is a test of installing openstack and then seeing if it works.
OpenStack components do not need test-requirements to operate,
that's why they are test-requirements.
Additionally, as we look forward to depsolver pip, this is going
to screw us because we don't apply constraints to linters, which
are expressed in - you guessed it, test-requirements.
Change-Id: I8f24b839bf42e2fb9803dc7df3a30ae20cf264eb
This reverts commit 79b8e79488.
This is breaking things in various jobs, most notably because
we do not put constraints on linters - but we install
test-requirements which then can conflict with each other.
Change-Id: Ibc5603c61b38ce44db58fb27a27352f59123ad09
The new pip depsolver is coming this summer. Until it's ready,
run pip check at the end of devstack to make sure we're not
somehow installing conflicting package versions. We shouldn't
be, because of constraints, but if we are, better to know and
start figuring it out.
Change-Id: Id98f0848ff5a252d93e5f8029df2d069924d603f
The function was using a hard coded value of localhost:11211 when
we have an option MEMCACHE_SERVERS that can be defined and used
inside DevStack.
Change-Id: I4947928fe406a9844d5bdaa3c826d273952fa097
DEVSTACK_GATE_FEATURE_MATRIX seems to be an old legacy thing that
is no longer being used. It currently prevents using the jobs in
openstack/devstack without adding openstack/devstack-gate for the
role.
Change-Id: Iab9b4862c01043d2c158398bac4b3b289a0adba0
tempest-multinode-full is py2 job and not needed to
run on ussuri onwards. Chaning this to its py3
version tempest-multinode-full-py3
Change-Id: Iff271eabcf1a39d6bf6c1fcd55ff2749cab2373f
The networking-ovn code is being moved into the neutron repository as
part of the effort [0].
This change is needed so we are able to install OVN from packages
when running the networking-ovn functional tests along with
the Neutron ones (see [1]). In the old networking-ovn repository we did
compile OVN from source instead of installing it from packages but
that took time. We want to do better in the Neutron repository.
[0] https://blueprints.launchpad.net/neutron/+spec/neutron-ovn-merge
[1] https://review.opendev.org/#/c/697440/
Change-Id: I92ab727d9954eb729c41b9a67ecb60b56883097b
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
From Train release Glance has added support [0][1] to configure multiple stores
of same or different types. This patch enables developers to configure
multiple file stores for glance. In order to configure multiple file stores
user need to set below options in local.conf
GLANCE_ENABLE_MULTIPLE_STORES=True/False
To enable multiple stores of glance.
GLANCE_MULTIPLE_FILE_STORES=veryfast,fast,cheap,verycheap,slow,veryslow
Comma separated list of store identifiers.
GLANCE_DEFAULT_BACKEND=fast
Default glance store in which image should be stored if store identifier not
specified explicilty. Should be one of the store identifier from
GLANCE_MULTIPLE_FILE_STORES config option.
NOTE: This support is added so that we can start adding tempest/CI tests for
glance multiple stores.
[0] 515412b59f5b3af07a1787b9f8e85a4d656d3e1c
[1] https://docs.openstack.org/glance/train/admin/multistores.html
Change-Id: I494f77555cfe9115356ce0ee75c7d7f192141447
Building on I5c3e1b7b632fd73310c462530990cdb0e0c0ceea we can now add a
Fedora job using the virt-preview repo that will be used by Nova's
experimental queue.
Change-Id: Iad9d64912bb07f307e4897ece1621f275f1d5211
'str' type in python2 is 'bytes' type in python3,
when use python3, we will get a prefix 'b':
sudo ip netns exec b'qrouter-39cc2b45-a27c-49c5-94a6-26443a49ac63' ip
neigh
--------------------------------------------------------------------------
*** Failed to run 'sudo ip netns exec
b'qrouter-39cc2b45-a27c-49c5-94a6-26443a49ac63' ip neigh': Command 'sudo
ip netns exec b'qrouter-39cc2b45-a27c-49c5-94a6-26443a49ac63' ip neigh'
returned non-zero exit status 1.
The message above is raised by running tools/worlddump.py with python3.
Change-Id: Ic254af86fa27729839f00c0ad4a5bbbc9e545a09
With fedora-latest now using Fedora 30 images and Fedora 29 itself EOL
we can now remove support for it from devstack. This change also cleans
up references to earlier Fedora releases under files/rpms/.
Change-Id: I24332f7016ebb549ea678acf677c477b55ec4d4b
With the open-iscsi rebase we should now be able to move the
fedora-latest job to fedora 30 ahead of the removal of support for 29.
Change-Id: Ia80f03f275e595d3b26b52b3478f303036d59438
I2f16658c5a3e22cac70912a0f3ad65cdd7071a1e worked around an open-iscsi
bug that remains unpatched in Fedora 30 and 31 by using a private copr
repo with the patch applied. Rawhide has finally been rebased to 2.1.0
where this issue and many others have been resolved.
We can now use a new repo that has been built for Fedora 30 and 31 that
provides this rebased package that we can either use until the rebase is
backported to 31 and 30 or 32 is supported.
Change-Id: I5ba5885bd9c784949602aeb4ddff9d75fecc6b3d
With the recent switch to py3 we need to ensure python3-devel is
installed on Fedora >= 29 before pip installing any dependencies.
Change-Id: I21417687db5b3827c5272ed22f6dc95db13f2870
The GRANT command in mysql8 can no longer create a user implicitly.
Split that part into a dedicated CREATE USER command.
Also drop disabling the query_cache, it is off by default for some time
and the option got removed in mysql8.
Change-Id: I31bcc285ff8e373abbacb303c1269857c9cfa9ed
After Python 2 is getting unsupported, new distros
like CentOS 8 and RHEL8 have stopped providing 'python'
package forcing user to decide which alternative to
use by installing 'python2' or 'python3.x' package
and then setting python alternative.
This change is intended to make using python3 command as
much as possible and use it as default 'python' alternative
where needed.
The final goals motivating this change are:
- stop using python2 as much as possible
- help adding support for CentOS 8 and RHEL8
Change-Id: I1e90db987c0bfa6206c211e066be03ea8738ad3f
Set default python3 version as 3 for cases python3
is not installed before running DevStack.
Implements installation of required python3x package
for RedHat family distros with package name depending on
configurable ${PYTHON3_VERSION}. Examples:
3 => python3 python3-devel (default one)
3.6 => python36 python36-devel
3.7 => python37 python37-devel
This should help in situations where there are more
than one python available for given platform and
DevStack is asked to pick one by its full 3.x version
Change-Id: I49d86bc9193165e0a41e8e8720be37d81a4e7ee0
We have dropped the tempest-full from
Tempest gate and made this to run on py2
explicitly which is nothing but for stable gates.
- I75868d5c9b6630fe78958ff89e58a0aced09a6b3
This job is not supposed to run on ussuri onwards master
gate as everything will be python3-only.
Change-Id: I372bde6a1753884efaf15da5fab48f1bddb4dab5
We are hitting this error:
+ tools/fixup_stuff.sh:fixup_ubuntu:82 :
sudo rm -f /usr/lib/python3/dist-packages/httplib2-0.11.3.egg-info
rm: cannot remove
'/usr/lib/python3/dist-packages/httplib2-0.11.3.egg-info': Is a directory
This patch adds the -r option to allow removing folders.
Change-Id: Ib7bb8b0a3dcf747bcc06da1a2fb17fa9d8808484
When starting 'memory_peak' service is using python command instead of
python3, while psutil (required package) is most probably being
installed into the python3 environment (as we are dropping python2.7
support).
Closes-Bug: #1860753
Change-Id: Ia2b7e2e33d784560443131e2965f520b361a54e3
TEMPEST_BRANCH which is mostly set as master so
that Tempest master is run to test the env. With
stable branch going to EM state and Tempest master
might not work due to incompatibility of code or
requirements. In that case we pin the Tempest so that
older Tempest can be used for their testing.
Till now for ocata, pike and, queens we used the gerrit style
ref to pin the Tempest which is not preferred way. We should be
able to use the tag name on TEMPEST_BRANCH.
This commit explicitly checkout the tag set in TEMPEST_BRANCH
as git_clone does not checkout the tag directly until RECLONE
is true or tempest dir does not exist.
After this stable branch or job can set the tag directly with name.
For exmaple: TEMPEST_BRANCH=23.0.0.
Change-Id: Ic777e4b56c4932dde135ac909cb5c6f4a7d5cc78
Work on the new neutron scripts has stalled and they aren't in a useable
state yet. Given the ongoing decline in contributions, let us
acknowledge this and undeprecate the neutron-legacy scripts so that
people can continue to use them without feeling guilty about it.
Change-Id: I4bce19da861abf18ddb89d82fd312c5e49a4ee7c
The virt-preview repo provides the latest rawhide versions of QEMU,
Libvirt and other virt tools for older releases of Fedora. This repo is
extremely useful when testing features in OpenStack that rely on these
latest builds well in advance of them landing in full Fedora, CentOS or
RHEL releases.
This change adds a ``ENABLE_FEDORA_VIRT_PREVIEW_REPO`` configurable
to control when this repo is enabled and used when deploying on Fedora.
Change-Id: I5c3e1b7b632fd73310c462530990cdb0e0c0ceea
'tools/mlock_report.py' script requires 'psutil' package to be
installed. This ensures it is done before memory_peak service is
started.
Partial-Bug: #1860753
Change-Id: I7b2b6eaf9856c6057e1a4a0054d15074150a6cb6
The hack has be around for pip 1.4.1 and older. It should be safe to
remove it by now. In fact it causes problems in my Ubutu Bionic VM when
trying to overwrite httplib2 library installed from the distro package.
Change-Id: I34b826f4e8f10f8d44b888120f19fcc7ba501b3d
[1] stopped installing pip for py2 when py3 is being used.
This patch makes sure we check only for py3 pip then.
Also removed some no-longer-relevant comment and
made uninstall behave the same.
Check for pip>=6 removed too.
See also [2].
[1] 279a7589b0
[2] http://lists.openstack.org/pipermail/openstack-discuss/2020-January/012182.html
Change-Id: I36ee53e57e468d760b80a7e621b90899867a8efd
We dropped the ability to override USE_SYSTEMD to False when we deleted
screen support in [0], so we can also clean up any conditionals based
on it.
Also clean up the logging setup functions, dropping local vars for
parameters that we don't actually support anymore.
[0] I8c27182f60b0f5310b3a8bf5feb02beb7ffbb26a
Change-Id: I5cbce9f2c42e111761e8689447b3f8cbb7ea2eb5
The Sheepdog project has been defunct for awhile now, and the Sheepdog
driver and os-brick connector is now being removed from Cinder. This
cleans up plugin references for the driver.
Change-Id: Ieb2d9cf653b2d3a4af30cab26b8428a7c7edff98
Signed-off-by: Sean McGinnis <sean.mcginnis@gmail.com>
The proposal job to update the plugin list has been failing for a long
time as it gets a 500 error from gitea on the openstack/openstack
repo. This is an odd "superrepo" with all projects as submodules;
thus openstack/openstack/devstack is actually a project, not the
directory with a plugin in it.
Skip this repo (gitea shouldn't return a 500, but that's another
thing...)
Regenerate the list manually for this run.
Change-Id: I6ed65bcb720d4cb10702cbf66106120e001ec35f
Some distros do not install python3/python3-devel with the minimal
install. F29 doesn't install -devel, and neither Centos 7 or 8
install either. This patch ensures that these packages get installed.
Ideally, PYTHON3_VERSION would be set *after* ensuring that python3
was installed, but it gets a little tricky with all of the includes.
This sets it to 3.6 as nothing uses 3.5 anymore.
Change-Id: I7bdfc408b7c18273639ec26eade475856ac43593
Need to make PyYAML overridable on Ubuntu, it is a dependency for e.g.
cloud-init, so we cannot remove it.
Depends-On: https://review.opendev.org/703792
Change-Id: I4423dfb2c30299903b52a2bb06d846dd487f5b8b
nova-live-migration is legacy job and and rely on
devstack-gate + devstack setting so any change in devstack can
break it. Example bug: 1860021
We can remove this job once it is migrated to zuulv3 native.
Change-Id: Ie34d4dc1ab30ced8161796fe32628db07de86cc9
Related-bug: #1860021
In same cases, the hypervisor presents to the guest OS a named CPU model
is similar to the host CPU and adds extra features to approximate the
host model. However, this does not guarantee all features will be
precisely match.
This patch adds LIBVIRT_CPU_MODE to allow users to define the CPU mode
they want to use, for example "host-passthrough".
Change-Id: I83792c776b50d1d22584be2a37cc6a166f09c72b
This reverts commit d8dec362ba. This has
knock on effects for devstack-gate, which configures g-api on subnodes
node but not mysql, resulting in failures. A longer term fix would be to
either a) stop configuring g-api on subnodes if we can determine it's
not necessary or b) only configure the database if on the main node.
However, both options are subject to debate so for now just unclog the
gate.
Change-Id: I58baa3b6c63c648836ae8152c2d6d7ceff11a388
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Closes-bug: #1860021
This reverts commit d7dfcdb467. A
subsquent change that depends on this,
d8dec362ba, has knock on effects for
devstack-gate and needs to be reverted. Revert this first.
Change-Id: Ic5402f57052648e10eacf3c3de67d2cdd2d42f63
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Partial-bug: #1860021
Everything in OpenStack *must* be Python 3 supporting now, which means
it's time to remove the functionality that allows us to blacklist
packages that didn't support Python 3.
Change-Id: I7c8cf538ec88bd4056b0109f19671e3d65f5da3a
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Since [1] Glance init depends on either g-api or g-reg being
enabled.
This broke multinode g-api deployments with singlenode database
backend.
This commit aligns Glance with other services w.r.t when to
apply database init.
[1] d8dec362ba
Change-Id: Idc07764d6ba3a828f19691f56c73cbe9179c2673
Closes-bug: #1860021
This has now been removed and even prior to removal defaulted to True.
Change-Id: I847a873d833a4dbee96afa1d2726fea2b8045eeb
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
This option has default to the 'NoopFirewallDriver' for some time and
will soon be removed. Stop configuring it entirely.
Change-Id: I4dbc0015cf26d7edf51d0d5fd978ccd3a1ad1b79
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Currently VMs created by a previous DevStack run still exists even
after re-run stack.sh. This leads to a failure in launching a VM after
the second run of stack.sh. We need to check the reason of the failure
by nova compute log and clean up remaining VMs. It is annoying.
IIRC we cleaned up existing VMs. While I failed to identify which commit
changed this behavior, I believe it is worth recovering it.
This commit changes unstack.sh to call cleanup_nova. cleanup_cinder()
already cleans up LVM volumes and some of them may be used by VMs,
so I believe it is reasonable to clean up VMs in unstack.sh.
Change-Id: I9fcbc5105e443037fada1ef6a76a078145964256
This change removes the .gz extension from the
service and syslog logs exported via journalctl.
This change nolonger gzip compresses the exported
logs so that they can be rendered in the browser
directly when served from swift.
Change-Id: I4557a737cb13b9c2406056be08ab8a32ddd45162
This was added in 2012 with I89677fd54635e82b10ab674ddeb9ffb3f1a755f0,
but I can not see it being used anywhere currently.
It's use of virtualenv's has become problematic in a python2
deprecated world, but since it is not used, remove it to avoid further
confusion.
Change-Id: I65d44d24f449436ca6229928eee2c5a021793055
Just calling "virtualenv" makes a Python 2 based environment;
setuptools just dropped Python 2 support (as Python 2 reached EOL in
Jan 2020) so this has now become a breakage.
Although the Python 2 path won't work, use the abstracted command.
This should stop us having to revisit this for any future cleanups (or
switing to venv, etc).
Change-Id: I531e971b78491a9276753c0d86b04c4adbd224aa
In Train, the access rules API was introduced in keystone. This change
enables testing it in tempest.
Depends-on: https://review.opendev.org/699519
Change-Id: I2af21868cbf584a6881c6208bc2afc3bdb323ab9
lib/tempest uses 'tempest-account-generator'
which is deprecated 4 years back.
In addition to above, lib/tempest also uses
'os-tenant-name' which is also deprecated.
Use of 'tempest account-generator' and
'os-project-name' should be done now.
Signed-off by: Soniya Vyas<svyas@redhat.com>
Change-Id: I624e1dc57a3d3533322fb298c01f70241d0400ed
Tempest's service_availability config option includes all the service
availability which is further used by tests to take decision of skip
or run the test.
For example, [service_availability].nova is true then, compute test will run
or if [service_availability].aodh is false then, all aodh related tests either
in aodh tempest plugin or any other plugins will be skipped.
Now question is what is the best way to set the each service availability for
tempest or tempest plugins tests. We have 2 category of service here-
1. Service tested by Tempest (nova, cinder, keystone, glance, swift, neutron)
(let's say type1 service)
2. Services tested by Tempest plugins (all other than above list)
(let's say type2 service)
We need the standard way to set both type of service so that we can maintain
the setting of service_availability config options in consistent way.
As discussed on bug#1743688/ and review https://review.openstack.org/#/c/536723/,
we will use devstack lib/tempest to set the type1 service which is services test
owned by Tempest and type2 service setting will be done by devstack plugins of
those service.
For example - [service_availability].ironic will be set by ironic's devstack plugin.
because that is best place we know ironic is installed and available.
To do that we need:
1. Add setting of [service_availability].* in devstack plugins
2. Remove setting of type2 service from devstack lib/tempest
This commit does the second part and all depends-on patches handle the first part.
Related-Bug: #1743688
Change-Id: If3aec9fd1c61e2bb53233be437b97b811dc82414
This package provides dhcp_release tool but in files/debs/neutron-common
it was listed to be installed only on Ubuntu Precise.
The same file is also in Nova's packages but there is no restriction to
Ubuntu Precise only there.
So on all Neutron jobs it was fine but on Ironic's job where Nova
wasn't enabled, this package was not installed and caused problems
in Neutron DHCP agent.
Change-Id: Idd0711cfe6d43f21754a2f0c230cd094ea33cb27
Closes-Bug: #1855910
We've seen jobs where tests fail due to what appears to be rng
starvation. Enable virtio rng device to try and alleviate this.
Change-Id: I70d800cdc45b6008f775110f22c0000736421529
input() should work on both python versions for what we need. I
understand the concern about eval() on python2 but, in the case it's
used we should be fine, plus, python2 is being removed from OpenStack
projects.
Change-Id: I86a7c31374986f81132bc4f49aee0a76b90e6553
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
This variable can be now set in Devstack's config file and in
such case Devstack will not set it automatically to value most
likely correct for the distro.
By default this value is empty string and in such case Devstack
will work in exactly same way as it was before this patch and
will determine automatically what name should be used there.
In addition in case of Ubuntu package $MYSQL_SERVICE_NAME-server
will be now installed instead of mysql-server always.
This will allow to easy configure e.g. CI job which will run using
Mariadb instead of Mysql on Ubuntu.
Change-Id: I25af0b54ad235b08c6c399b4125c737acf57ee2e
This will avoid the creation of an unneeded file in the "tempest"
repository directory.
TrivialFix
Change-Id: Id3f46b3537cd3232cb29c42808bde44c667565f1
With the goals of Ussuri being Python 3.6 [1], the python 3.5
environment on Xenial is too old. Remove testing and the most obvious
bits of support from devstack.
Also drop claimed support for artful, which is long EOL.
[1] https://governance.openstack.org/tc/reference/runtimes/ussuri.html
Change-Id: Iefcca99904dde76b34efbbfc0e04515dfa5a09e5
This reverts commit f99d1771ba.
Added workarounds that might want to get split into their own patch
before merging:
- Don't install python-psutil
- Don't run peakmem_tracker
Change-Id: If4fb16555e15082a4d97cffdf3cfa608a682997d
Since Stein, gate jobs have been using bionic nodes so they
are running with python 3.6, so it makes sense to also default
devstack itself to run with python3 by default.
Depends-On: https://review.opendev.org/688731
Change-Id: I52b03caee0ba700da3a15035201ea6cd91baa06b
We need to source the environment overrides before they get evaluated.
Otherwise e.g. USE_PYTHON3 is factually being ignored for some settings.
Also fix creating python3 venvs by using the "virtualenv" command for
that task.
Change-Id: I16c78a7fef80372d9a1684c3256c5b50b052ecae
This job is still running python 2.7. As we are dropping py2 support in
Ussuri cycle, lets drop this job now.
There is same job called "grenade-py3" which runs on python 3 already
and this will be now used in project's CI.
Depends-On: https://review.opendev.org/#/c/695036/
Change-Id: I5cd8e137a3ae06e49a4351629c5eb207c4e6bf1a
The centos7 job is running with python2, which is no longer supported by
nova, so we can drop it in master.
Change-Id: Id9ef507dd6f4226d65c6ed3043666b0aa6a3bd1c
With Glance defining default policies in code, it's no longer necessary
to install policy.json from the repo.
Change-Id: I9f9160f5a2bf9fd77fb3807e12de219b7a49952d
Depends-On: https://review.opendev.org/693129
In cases where global REQUIREMENTS_DIR is set, use it
instead of overwriting it. This is particularly needed
in cases where users of pip_install wrapper have the
upper-constraints.txt at another location.
Change-Id: I34e9f94548c575e1af5bca9655a3b7d1915375a8
Signed-off-by: Flavio Fernandes <flaviof@redhat.com>
Some distros no longer ship brctl, iproute2 should be used
in its place. The linuxbridge agent plugin script was still
using it, as was worlddump, which generates this warning on
a failure:
Running devstack worlddump.py
/bin/sh: 1: brctl: not found
Conditionalizing worlddump based on whether brctl is installed
to make this go away.
Change-Id: Iafbf4038bab08c261d45d117b12d4629ba32d65e
When configuring TLS between the console proxy (where the n-novnc
service runs) and the compute host, some configuration for QEMU needs
to be done on the compute host. The existing code for this requires the
n-novnc service to be running, which it is in a single node all-in-one
deployment. However, when running in a multinode deployment, the
n-novnc service runs only on the controller and not on the subnode.
Yet, we need to configure QEMU on the subnode compute host as well.
This removes the n-novnc service requirement to enable TLS QEMU
configuration to occur on a compute subnode in a multinode deployment.
Closes-Bug: #1849418
Change-Id: I8b6970e91ad7f52ff489cb9f776ca216d8f86aa4
Based on resolution [1], there's no clear indication that next
steps involve the removal of the DB from Devstack or from the gate.
[1] I332cef8ec4539520adcf37c6d2ea11488289fcfd
This reverts commit d9aaae95f2.
Change-Id: I8410d65c0e0b24035aa035fac7560a686d53ec50
This reverts commit 168ca7f0a4.
Removing postgresql support from devstack was unnecessary
since it's not broken and not causing maintenance issues
as far as I know. The commit being reverted said that pg
support was deprecated in Pike but nothing in the docs or
commit message refer to official deprecation of postgres
support in devstack or openstack in general. Not to mention
that there are still postgres-based jobs that will no
longer work *and* the notification to the mailing list about
doing this happened *after* it was already done [1] leaving
stakeholders with no time to reply.
[1] http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010196.html
Change-Id: Ie7036d37d79e6aba462b7c97f917e2e7aed108f9
This only applies to the LVM driver (when using
thick provisioning), and doesn't have any effect on
other backends like NFS, so only write the conf entry
for LVM.
Change-Id: I722ba2fa0010d9887ed9b7fdd9e050cd4694768e
keystone v3 is the default API version in horizon now,
so there is no need to specify it in local_settings.py explicitly.
This commit also makes minor changes in lib/horizon
_horizon_config_set().
* Do not insert a blank line after each setting.
* Use the local variable $file to specify the target file
consistently.
Change-Id: I5faea3e1f357726a256d2b48fc1afeabfead4998
The default permissions for the zuul home directory
are not the same in the various distributions.
As /home/zuul contains the sources, a 700 default may be
problematic when accessing those files, so make sure
that the executable permissions are set.
Closes-Bug: 1846251
Change-Id: Ic9769e56274d7205844b86d3b5200a6415e4acad
Since Queens [1] nova has been able to be configured with
cinder service user credentials for operating on cinder
resources without a user auth token similar to things nova
needs to do without a user auth token for working with neutron
and placement resources.
This change:
- centralizes the nova [cinder] section configuration
- adds the necessary auth configuration
Needed by: https://review.opendev.org/549130/
[1] I3c35bba43fee81baebe8261f546c1424ce3a3383
Change-Id: I5640ee431f6856853f6b00ec7ed1ea21d05117dd
openSUE Leap 15.0 and 15.1 both provide python3-six version 1.11.0.
Since version 1.12.0 was released, pip>=10 recognizes the version
difference and tries to uninstall the distro-provided version and fails.
This change adds another hack to remove the egg-info file for the six
library so that pip can manage it directly. We also have to wait to
install os-testr until after the fixup has happened since trying to
install it triggers the issue.
Change-Id: I4649abe06b5893a5251bfcdd4234abccde6ceda2
The target branch was centralized in change
I82aa19e739eeda3721bac1cb5153ad0bf2d1125a but there
are two issues, pbr and diskimage-builder are using
TARGET_BRANCH which gets changed to stable/* for
each openstack stable branch that gets created for
devstack, e.g. I861068ae1a9902cef61c52c70dda7bb42f4371a0,
but pbr and diskimage-builder don't have stable branches
so they should be using BRANCHLESS_TARGET_BRANCH i.e. master.
Change-Id: I47ac7a7e194ca6d613d0ccaebfd557346644c2df
A 500 error from gitea can occasionally show up as a project dropping
their devstack plugin (I543faced83a685d48706d004ae49800abfb89dc5).
To avoid noise in the proposal jobs, implement a small retry loop for
500 errors.
Change-Id: Ide23e4de819a2c751d887eeaa7f0b9d0437f8e2c
I forgot in I2f16658c5a3e22cac70912a0f3ad65cdd7071a1e that "is_fedora"
matches rhel/centos for historical reasons. Restrict the install to
just the Fedora platforms by checking DISTRO matches
Change-Id: Ica4a690a4f2894a03ceb8557a947ed2ea4a60e53
We need to enable accept_ra before we enable forwarding, otherwise
existing addresses and routes may get dropped until the next RA is
received, possibly causing connection errors in the meantime.
Change-Id: I1fdeede59547de896ed89222ecf121fd9e6b810d
In Fedora 29 dstat was merged with pcp-system-tools (see
https://pagure.io/fesco/issue/1956)
Work around a iscsi bug with external packages until we can get the
package fixed.
Obsolete F27
Change-Id: I2f16658c5a3e22cac70912a0f3ad65cdd7071a1e
All credit for figuring this out goes to frickler (and that was the hard
bit so thank you!). The worlddump files were not being collected because
they weren't in our log collection list. Add worlddump to this list
so that we collect these files.
One thing that makes this slightly complicated is the worlddump files
are named with a timestamp and we can't have globs in our collection
list. To address this we create a copy of the file with a -latest.txt
suffix. This gives us a deterministic file name for log collection
without using globs.
Note we do not use a symlink here because some jobs gzip their log files
(breaking symlinks) and others do not. This makes it painful to always
have a valid link. Not having a valid link can break log collection.
Hardlinks may be another option but simply making a copy is easier to
manage as you don't have to worry about links preexisting and the
dumpfiles are not that large.
Change-Id: I96ae5f5290546ad25ca434c1106c01354d2d053c
All the OpenStack projects should be able to run under Python 3 now so
the fallback installation of the Python 2 libraries should not be
needed any longer. This also avoids the problem of script files
installed by the libraries sometimes being overwritten by the Python 2
version leading to incorrect execution later, as discussed in
http://lists.openstack.org/pipermail/openstack-discuss/2019-September/009226.html
This reverts commit a2eb89417f.
Change-Id: I1cdb7e4a209872f1620be556b7278879a4b86df5
The script currently complains about multiple default security groups. This
obtains the default and uses it when creating security group rules.
Change-Id: I81e59eae5df79889ed1fb02d45af26e3a55aa0e9
This was deprecated for removal in Pike. It's probably time to drop it.
Note that the 'postgresql-devel'/'postgresql-server-dev-all' packages
are retained since some packages still include 'psycopg2' in their
general requirements.
Change-Id: I51e8354e99972757253ce259e6c03c91da24398c
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Currently if user selects the default instance type for Tempest
tests, some of resize-related tests may fail due to resize attempt
into flavor with smaller disk size. It is because there is just
simple check if flavor_ref and flavor_ref_alt (IDs) aren't the same.
To ensure resize is really possible, there shall be additional
verification introduced.
Co-Authored-By: Michał Madarasz <michal.madarasz@corp.ovh.com>
Change-Id: Iaa1bfa9cb76cbe54be658d2d70d97d99e7fb5be9
Per the referenced bug, $NOVA_CPU_CONF was previously being initialized
by copying $NOVA_CONF, thereby trashing any values already configured in
$NOVA_CPU_CONF.
With this commit, we merge the values from $NOVA_CPU_CONF in after the
copy.
Note that this makes use of the merge_config_file function, which is
defined in inc/meta-config, which wasn't being sourced from every code
path that hit start_nova_compute; so this commit also moves that import
from stack.sh to functions (next to the other imports from inc/, which
makes sense anyway).
Change-Id: Id3e2baa2221e13f512f8dcf1248e1e15b6a7597f
Closes-Bug: #1802143
We're trying to get nova to talk to ironic through openstacksdk and need
to be able to specify retry limits/intervals there. We could reuse the
existing conf options, but better to support the standard ones exposed
from keystoneauth1 via [1] and [2].
Note that these will be ignored unless you have keystoneauth1 3.15.0
(for [1]) or 3.16.0 ([1] and [2]) and are building your adapter using
ksa-derived conf options (see the Needed-By).
Needed-By: https://review.opendev.org/642899
[1] https://review.opendev.org/#/c/666287/
[2] https://review.opendev.org/#/c/672930/
Change-Id: I79c416e25d635b0ffa419640b4bd91e36f78b1ab
Use the PUBLIC_NETWORK_NAME variable instead of hardcoding
it when setting the [ml2_type_flat]/flat_networks option.
Change-Id: I8bfc37089ec90eb06ee41d85744dad0f3f734c16
Devstack's lib/cinder set the my_ip on cinder side but it
hard-code it with HOST_IP[1]. It is no issue for IPv4 env
but when you build or run the IPv6 job then this ip is left
to set with IPv6. my_ip should be set to HOST_IP or HOST_IPV6
based on SERVICE_IP_VERSION value.
As part of Train community goal 'Support IPv6-Only Deployments',
we will expand the 'devstack-tempest-ipv6' job to do IPv6-only
deployments verification so we need fix the the my_ip setting.
Closes-Bug: #1838250
Depends-On: https://review.opendev.org/#/c/677524/
[1]https://github.com/openstack/devstack/blob/6aeaceb0c4ef078d028fb6605cac2a37444097d8/lib/cinder#L231
Change-Id: I71c74e46467a5d3c1bf9c7d683f364cba7cf9d80
F28 is EOL as of 5.29.19, so change to use fedora-29
and remove it from stack.sh.
Depends-on: https://review.opendev.org/#/c/662538/
Change-Id: I5ebdb68fcd01a1e63be4b3c0735a274783aad818
Reformat overlong table into a list-table to make it easier to edit.
The change contains no wording changes besides giving titles to the
list-tables.
Fix formatting for setup-devstack-source-dirs so that the variable does
not get displayed with a grey bar at
https://docs.openstack.org/devstack/latest/zuul_roles.html#role-setup-devstack-source-dirs
Change-Id: I7378d46c507b1d86f1d5319655a55f2a8c5a8f60
For some crazy reason, we've forgotten about trying
to use IPv6 addresses directly with the SSL certificates.
So lets add some logic so clients can connect directly
with the v6 IP.
Change-Id: Ie8b8a2d99945f028bebe805b83bfd863b7b72d57
Switch to "modern" way of building docs using sphinx-build directly,
remove now unsed parts from setup.cfg.
Upgrade to openstackdocstheme 1.20 and remove obsolete variables from
conf.py.
Convert external links to internal RST links so that Sphinx can verify
that they are correct.
Replace redirected links with new targets.
Use opendev.org instead of github.com where appropriate.
Change-Id: Iedcc008b170821aa74acefc02ec6a243a0dc307c
Currently, the console server host and listen address on the compute
host is always being set to localhost. This works fine in a single
node all-in-one deployment, but will not work properly when
nova-compute is running on a separate host in a multi-node deployment.
This sets the console server host and listen address on the compute
host to the nova host IP and service listen address instead of the
localhost.
Co-Authored-By: Matt Riedemann <mriedem.os@gmail.com>
Closes-Bug: #1669468
Change-Id: Id8b0b4159b98c7ff3c85ec3daa03d556d9897ce9
A huge part of the logs is irrelevant bash aliases captured by the
openstack client timing wrapper from the run of "openstack complete",
which is only helpful on interactive systems where you'll interact
with the command line. Call it directly to avoid capturing the logs.
While we're here, turn off tracing inside the oscwrap function, which
is called frequently. It's not useful for debugging.
Change-Id: I1cb5399fe7ee6f0e547a9cfff70396aa2007632e
This patch adds new options:
* CACHE_BACKEND - with default "dogpile.cache.memcached"
* MEMCACHE_SERVERS - with default "localhost:1121"
to add possibility to configure various backends as cache in
Nova and Keystone.
It also adds options:
* KEYSTONE_ENABLE_CACHE - True by default
* NOVA_ENABLE_CACHE - True by default
To make possibility to enable and disable cache in those projects'
config files.
Default values configured there are the same as before were
hardcoded for Keystone config.
Nova has also enabled this cache by default.
Change-Id: I9082be077b59acd3a39910fa64e29147cb5c2dd7
Closes-Bug: #1836642
We recently added the 'tempest-ipv6-only' job on
tempest side which will use the devstack base job
'devstack-ipv6' job and add more verification and testing
for IPv6 deployment.
Let's add that job on devstack gate also to avoid any
break due to devstack changes.
Change-Id: Ib2c85ec262b027351872e2b5a39b06a4ba1b880a
Story: #2005477
Task: #35923
In lib/neutron-legacy, the Nova metadata host address is
un-quoted if it is IPv6, i.e. 2001:db8::1, not [2001:db8::1].
We should be doing the same in lib/neutron.
Change-Id: I80c96603a41ef9d289712ef15b464859aa9257be
The api documentation is now published on docs.openstack.org instead
of developer.openstack.org. Update all links that are changed to the
new location.
Note that redirects will be set up as well but let's point now to the
new location.
For details, see:
http://lists.openstack.org/pipermail/openstack-discuss/2019-July/007828.html
Change-Id: I8a6c3403192d1416cb66cc9e92ec827b339f1270
* The restart loop for rabbitmq-server can trigger socket activation
of epmd without rabbitmq-server running. This can lead to 'systemctl
status' reporting 'State: degraded' with no simple way to reset to
'State: running'.
* It's important to note that this socket activation failure is benign
and is not an indicator of system failure.
Change-Id: Iede4f5ebeffb59644dee4a17b6331b3cdd04d146
Signed-off-by: Jan Gutter <jan.gutter@netronome.com>
running chown and chmod on files and folders not created by
devstack causes a few issues:
* On nfs mounted directories it can take an extremely
long time to chown -R some of the git repos, especially
if any tox commands have been ran in the host
* chown can cause the host files to get into a weird state
if nfs is set up wrong.
If files and folders are pre-existing we should assume
they are in the correct state, and not modify them.
Fix setup-devstack-log-dir to create the logs directory with
correct permissions in the first place.
Change-Id: I5ebdaded3ffd0a5bc70c5e9ab5b18daefb358f58
Signed-off-by: Graham Hayes <gr@ham.ie>
'devstack' job set the VNC listen addresses 'VNCSERVER_LISTEN' and
'VNCSERVER_PROXYCLIENT_ADDRESS' IPv4 which makes 'devstack-ipv6' job
to either unset those or set for IPv6 values.
Let's remove the setting of those in base job and let lib/nova
set based on configured ip version from job.
'devstack-ipv6' base job will be used to define the IPv6-only jobs
on Tempest and project side gate.
Change-Id: Iea469128b15298aee61245e702d20603c8d376fb
Story: #2005477
Task: #35923
'devstack-ipv6' job set the devstack to deploy
the services on IPv6. As part of community goal
'Support IPv6-Only Deployments'[1], this is going
to be the base job for all project specific or tempest
IPv6 jobs. Running this as voting make sure any devstack
setting or changes would not break the IPv6 jobs.
Story: #2005477
Task: #35923
[1] https://storyboard.openstack.org/#!/story/2005477
Change-Id: Id6580e8b29b6b04e34c2c1eca3125fa08920eb1d
This adds a variable to control the [DEFAULT]/shutdown_timeout
config in nova to control whether or not a guest should have
a graceful shutdown of the OS or if it should just stop
immediately (no timeout). Since devstack uses CirrOS images
by default, the default value for the NOVA_SHUTDOWN_TIMEOUT
variable is 0 which should speed up tempest runs. The default
in nova.conf [1] is 60 seconds.
[1] https://docs.openstack.org/nova/latest/configuration/config.html#DEFAULT.shutdown_timeout
Change-Id: Ida83f70a1c4e61e5248f2bd42b4c24f7ac6d2310
Related-Bug: #1829896
Trying to deploy OpenStack environment consisting of ironic nova
hypervisor & hardware Ironic node (not VM Ironic node) with devstack
got failed.
Devstack error says error occurred while calling configure_libvirt
in configure_nova_hypervisor. This happens because libvirt related
packages are not installed when specifying "VIRT_DRIVER=ironic"
and "IRONIC_IS_HARDWARE=True".
To fix this problem, this commit add "if" statement to check
Ironic node is hardware or not using "is_ironic_hardware" function
in "function-common" file.
Change-Id: I1113478175fadec79d0f8bf6ae842ed86e5e686b
Closes-Bug: #1834985
opensuse-150 nodeset is referring to openSUSE 15.0, which is still in
maintenance but openSUSE 15.1 has been released already. "opensuse-15"
is going to refer to the "latest openSUSE 15.x" build released and
working for OpenStack going forward, so add this nodeset and use
it by default going forward.
Change-Id: Ic3f4d6998a66da5226bc95088d7e3c83dfe737ce
The package has been renamed in order to avoid the namespace collision
with systemd, update our doc accordingly.
Change-Id: I1b9a434d9bb6a7d9dc38ef965017ed9f8773d595
Closes-Bug: 1825949
In change I8934d0b9392f2976347391c8a650ad260f337762, we began
configuring console proxy ports for multiple cells in the nova
controller config files to avoid "Address already in use" errors from
port collisions when running multiple cells on a single host.
This correspondingly configures the console proxy ports in the nova
compute config file based on what cell we're in, according to the
NOVA_CPU_CELL variable.
The base_url config for serial console is also added where the default
was previously used. The url is taken from the config option default in
the nova code: nova/conf/serial_console.py [1].
[1] https://github.com/openstack/nova/blob/8f00b5d/nova/conf/serial_console.py#L54
Change-Id: Id885fc5a769bce8111f1052a1b55d26be817c890
Closes-Bug: #1830417
The URL for rdo-release package is version-less and redirects to latest
stable version. This becomes problematic when stacking older stable
versions as dependencies might not be met or newer and incompatibile
ones might get installed.
Closes-Bug: #1833696
Change-Id: Icb07dcb4c9a3950a3c31a3a8dcb8d0b4c713fdb1
This is no longer being used due to Keystone PKI tokens no longer
being implemented.
In order to not break backward compatibility we create a new function
that is to be used instead and deprecate the old one. Modify the old
function to ignore the 3rd argument and display a deprecation warning.
Adjust callers to no longer create and set that directory, calling the
new function instead.
Change-Id: Id0dec1ba72467cce5cacfcfdb2bc0af2bd3a3610
Currently, devstack has NETWORK_API_EXTENSIONS var to define
the network API extensions. NETWORK_API_EXTENSIONS is defaulted
to 'all' for master and hard coded list of extensions per release.
Zuul jobs of network extensions (for example neutron-fwaas) need
add the some extra extensions in the default list. To do so, they
need to duplicate all the defaults extensions and then add the extra
extensions. Much difficult situation is when defaults extensions list
vary from release to release so they have to keep updating the
NETWORK_API_EXTENSIONS per release.
This commit defines a new var ADDITIONAL_NETWORK_API_EXTENSIONS which
will take extra extensions and append into the default list. This way
Zuul jobs do not need to duplicate the default extensions.
Change-Id: I7270c9b9e047a851970439522c0356c9089a5b74
This is a follow up for change Ifcfce490edb3d77e4e436e002d35bc909e1a057c
where the GIT_BASE was changed to the opendev URL to avoid redirects.
The pbr namespace in stackrc was old and still getting redirected and
this change fixes that.
Change-Id: Ib444e928fa2ca7650670f97be6927202333a1dd7
This updates links going to git.openstack.org and review.openstack.org
to go to their respective opendev locations to avoid redirects.
Change-Id: I78e3bb5303718962f591117f9c0ee11f2314b128
Closes-Bug: #1833256
Update the server to opendev and update paths for gitea, along with
any other references.
Switch to a blacklist where we just remove stackforge; this leaves all
the new namespaces like x/ and starlingx/ being checked.
Use a common session for checking for the plugin file which makes it a
*lot* faster.
Remove unsed "plugins" array variable
Regenerate the file
Change-Id: Ie3e615ba352a389da22e129c5c67cf6abd8cfdc8
Many of the code blocks in the configuration documentation had extra
leading spaces. This resulted in the blocks being both code block
formatted as well as blockquoted in the output.
This patch removes leading spaces and some minor cleanup to get the
formatted output correct.
Change-Id: Ic4dfb49c547d51e16b673bc88d7b2b1a907e3258
Signed-off-by: Sean McGinnis <sean.mcginnis@gmail.com>
Since Ic0a03e89903bf925638fa26cca3dac7db710dca3 RetryFilter has been
deprecated. So we should not enable the RetryFilter on our tests.
Change-Id: I48c2c4d0714f582af8948dc88b48df1c2c62fcd2
Nova change https://review.opendev.org/603079/ changed the
default configuration to send only unversioned notfications rather
than both versioned and unversioned notifications. This could break
unsuspecting downstream projects (like Watcher) whose CI jobs are
not explicitly configuring nova for the types of notifications
they need but are just relying on getting both per the previous
default of the config option.
This adds a variable which defaults to "unversioned" to match the
nova default but allows downstream CI jobs to easily configure
another value.
Needed by https://review.opendev.org/663332/
Change-Id: Ied9d50b07c368d5c2be658c744f340a8d1ee41e0
Ironic's CI makes extensive use of VLAN based networking
and the newer neutron plugin hardcodes the tenant networking
type to vxlan which is naturally problematic. It also lacks the
ability to set the necessary constraints for vlan networking
which are added for vxlan networking.
This patch enables the type of tenant networking to be defined
as vlan, and enables for a physical network vlan range mapping
setting to be configured which is required for a vlan to be
allocated upon network creation.
Change-Id: I55874c1ce82898e9dfb81505d8f3b14abde33579
Nova has had a multi-cell job (nova-multi-cell) since the
Train release but is currently non-voting in the check queue
for nova changes. This change adds the job to the experimental
queue for devstack changes so we can test changes to devstack
and make sure they work for the multi-cell job.
Change-Id: Icf31baf6fd4313aec5ecfb9e8f9cbcef1ff7f61d
Tempest not support placement microversion setting so that
test can call APIs with specific placement microversion.
This commit adds the setting of placement API microversion on
Tempest conf.
Change-Id: Ie04aa993ec7a1495740d9267b076a40f4291e25e
If the sudo-ldap package is providing the sudo command instead of the plain
sudo package, accept that instead of breaking the system and requiring direct
root login intervention to fix things.
Change-Id: I45d7e4617bd59e72b4f0bf2e91750a6830e2a010
right now, there are duplicate titles under contents --> guides
as shown in https://docs.openstack.org/devstack/latest/, although
they will forward to different pages, it is still a little confusing
for users. This change will hide toctree in guides.rst and users could
click links in page to jump to detail page to read more.
Change-Id: I2f3abeca7f56a3aedeabb48630ed2c61635b93cd
There are still some issues with lib/neutron thus neutron-legacy is used
on controller node in multinode jobs and in single node jobs.
But in "group-vars" in devstack job it was configured to use
lib/neutron which can cause some problems in multinode jobs.
So lets switch to neutron-legacy on subnodes also until lib/neutron
will be ready to use everywhere.
Change-Id: I0d7f9f2baaee2836a719f199939156bd4f53f778
Change I188fc2cd1b26fe7a71804f7e7d66b111d6f15e30 in nova stopped us
respecting this when generating the network templates injected into
instances on boot. With the removal of nova-network, there is no longer
any other reason to set this.
Change-Id: I925b7c6c23133cd5a835960f4507c979f615d78e
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
In change I2ce8ff3d7c33a402b8af50182ec01f512859c388, we duplicated the
'default_floating_pool' option, found in the '[DEFAULT]' group, to the
'[neutron]' group. This allowed us to continue with our deprecation
plans for the former option, which should be retired along with
nova-network.
Update the nova lib module so it'll set the new option, we can safely
assume to be the correct one now that we've removed support for cells v1
and nova-network.
Change-Id: If9a02b640e6c2e1300c7b11b7552ba13c1496d79
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Remove nova cells v1 support, which also allows/necessitates removing
support for nova networks (which was only supported with cells v1) and
nova-consoleauth (which was required by cells v1 but is unnecessary
otherwise).
The Depends-On isn't really necessary, but it's here to make sure this
doesn't merge until we _really_ have killed cells v1.
I honestly expected this patch would be bigger.
Change-Id: I90316208d1af42c1659d3bee386f95e38aaf2c56
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Depends-On: Ib0e0b708c46e4330e51f8f8fdfbb02d45aaf0f44
We know empirically that some legacy gate jobs pass and appear
to be running with python3 but actually pip was invoked with
PYTHON3_VERSION unset so that they are actually ran with python2
packages.
As a followup to this discussion [1], add a safety check in the
get_pip_command function to ensure that a python version has been
set when it is invoked.
[1] https://review.openstack.org/#/c/622415/4/inc/python@283
Change-Id: I3a08406fb7d68282c6b98abb33a625821510046a
This patch enables ARP Responder with DVR routers along with
l2pop in devstack.
Related-Bug: #1774459
Change-Id: I82f628c32f6e38c2419b6ffe90d9f9adf96777b1
swift3 is no longer actively maintained in the upstream.
That has been moved to Swift repository as s3api so we should
use s3api middleware instead. As well as swift3, s3token is
also maintained in Swift upstream.
Change-Id: I4582d81da066ab53e6f11ad1df7af91425f2b0ca
Once nova is setup and n-cpu on the host is reporting
a service record and discovered (the host is mapping in
the API DB), we should run the nova-status upgrade check
to verify the deployment.
Change-Id: I9683bf94233ebacb3057ce159cb3dc53aa55a2f4
Related-Bug: #1790721
Option "lock_path" from group "DEFAULT" is deprecated. Use option
"lock_path" from group "oslo_concurrency".
Change-Id: I7c7501a4a351155eeba77bb7cd43c8d6f5ea73bc
Support for changing the cinder periodic_interval config option
was added way back in havana as a workaround for bug #1180976
by change I20e52e66fcc94b224476cdd14c88bd6981b4e617. As the fix
for that bug does not require modifying this config value, and
such modification may have unintentional adverse effects, end
the support.
Change-Id: I1ef1fe564123216b19582262726cdb1078b7650e
Partial-bug: #1824837
n-obj hasn't been around for many years and devstack
doesn't use it anymore anyway so this just cleans up
some vestigial use of the old service.
Change-Id: I04b2d2dc2b4e49fab90f5ef94f4e087e969aa24b
Tempest's scheduler_available_filters has a special 'all' value that
is understood to mean 'all filters are enabled' by various tempest
tests. However, what it really means is 'the default nova filters are
enabled.' In an effort to help clean that up, this patch explicitly
sets scheduler_available_filters to nova's $FILTERS. Because $FILTERS
is now used in both lib/nova and lib/tempest, it is renamed
$NOVA_FILTERS.
Change-Id: I6ffc1e9989cd61d666f9c1db9c94fbabd7151918
Related-bug: 1628443
With the new namespaces we have to look around a bit more to find
repos top copy into the DevStack working directory. Add:
* starlingx/
* x/
* zuul/
Depends-On: https://review.opendev.org/653988
Change-Id: I8a55522a5fee46f415f0c0ce580ded3476133460
Signed-off-by: Dean Troyer <dtroyer@gmail.com>
We should always pass on a region when talking to
ironic. This will also help detect and test issues
specific to regions.
Change-Id: Iaab3c1bcedc5aaa2106c0758cbb43bade3de2cf5
This function will now first filter out all "neutron-" strings from
DISABLED_SERVICES list before looking for "neutron" string in it.
Change-Id: I5cab6a3be553713e1257599fb72042c6001f2672
Close-Bug: #1824884
The python3-ply package is indirectly a dependency of dhcp-client, which
is not exactly an optional package. Pip >=10 refuses to install ply from
global-requirements with this distro package is installed, so our only
option is to remove it manually.
Change-Id: I377fdd4a581eb4b6275584d92cafc0b783fe3b84
openstacksdk gates on the new-style devstack functional base jobs. It
serves as a good test case to make sure the functional base jobs don't
break.
Change-Id: I817639ed30cda8ea51d156872a14bbcf10a4e63d
The pkg/elasticsearch.sh is only used by Panko but Panko has
moved the script to its own repository [1].
[1] https://review.openstack.org/#/c/643262/
Change-Id: I0ce40f4299246b68624abd2072c2abce06f1e70b
Signed-off-by: Trinh Nguyen <dangtrinhnt@gmail.com>
stack.sh usually fails when it is started in a shell session
where OpenStack related environment variables OS_* are set.
Most common failure scenarios are failures in keystone operations.
This commits clears OpenStack related environment variables
at the begining of stack.sh.
Change-Id: I3a924a0586dc9bb28f3bf3e151e100c24015efe5
Previously apache was configured and restarted before we configured
the CA and certs. In most cases this is fine because those specific
vhosts didn't use tls. However, if you had previously run devstack
and had leftover vhosts and an unconfigured CA or certs devstack would
fail.
This is a small corner case, but its simple to address by moving CA and
cert setup up in stack.sh to before we do anything related to web
servers.
Change-Id: I31dbaf9471088b9faff26c7b790da6f6feebb2d5
python2.7 will be EOL soon, let's test on python3 instead.
Exclude CentOS 7, as there is no python3 easily available.
Change-Id: I24d8812c0c37b6d376fd5ae38067513bb62a2804
We're able to run multiple cells in devstack by setting the variable
NOVA_NUM_CELLS in the devstack local.conf. Since we run console
proxies per cell, we will start two console proxies if
NOVA_NUM_CELLS=2. However, we've not been configuring the console
proxy ports in the nova_cellN.conf files, so an attempt to start
more than one will result in a port conflict and failure to start
the subsequent console proxy services with error:
ERROR nova error: [Errno 98] Address already in use
This adds configuration of the console proxy ports based on an offset
while looping across NOVA_NUM_CELLS. The base port values are taken
from the config option defaults in the nova code: nova/conf/vnc.py,
nova/conf/spice.py, and nova/conf/serial_console.py.
Closes-Bug: #1822873
Change-Id: I8934d0b9392f2976347391c8a650ad260f337762
When the role is used by grenade, the data directory is shared among
different devstack executions, and the base directory is different,
for example: /opt/stack/data vs /opt/stack/{old,new}.
The new devstack_data_base_dir parameter allows user to specify
a base directory for the data/ directory which is unrelated
to the devstack directory. The default value is devstack_base_dir,
so the default behavior is unchanged.
Change-Id: Ie69b7b51947cbf1a8b31d2701783de2fb56a2d33
If we are running with python3, just assume that any
package that is not blacklisted is available for py3
and just attempt to install it and let pip sort it out
whether it gets installed from a local or remote package.
Change-Id: Ic05d183e489320f6dfc721575d47e7e4d661f87c
Closes-Bug: #1820892
When [0] introduced quoting all arguments, it broke existing consumers
that already quote their value themselves. Fix this by avoiding to add
additional quotes to the value when it already starts with a double
quote.
[0] https://review.openstack.org/636078
Change-Id: I92146e04731efc6dcc632ae6c3a7c374e783cdba
Closes-Bug: 1822453
- Switch from proposing Ubuntu 16.04 to 18.04 as the most tested
platform.
- Make it clearer that creating an additional "stack" user is
optional when running on a cloud image, as this step often leads
to errors for new users.
- Fix some minor nits along the way.
Change-Id: I39aef1a230b668b932b1681fcd0deeb423b411f1
dstat is Python 2, never going to be updated and effectively
abandonded. The replacement is pcp-dstat [1] which is mostly
compatible, with a few differences. As distro start transitioning
(Fedora has), just drop the unsupported args for now.
[1] https://pcp.io/man/man1/pcp-dstat.1.html
Change-Id: Ibec8a37cb18a14656d97e2096c66bc8b21406068
This adds a service to run a tcpdump during the run. This can be
useful to capture various network traffic for post analysis.
There didn't seem to quite be an appropriate place to document it, so
a new debugging file is started, with some terse explaination of our
various system-wide debugging services.
Change-Id: I09aaa57611c5047d09a9bce7932d34e9d50b30e6
Currently we only export the devstack@ services, and then separately
export the kernel & sudo logs to syslog.txt.
This leaves a lot of logs potentially behind in the journal for
various daemons. Just export the whole lot.
Using this output is currently very opaque and makes use of systemd
export tools that are very un-discoverable. Add a README that will
appear alongside the journal explaining how to actually use it. This
is a template as it would be nice to put into things like the list of
services that are in the journal, or maybe other magic.
Also make sure we export the logs since the start timestamp; currently
during a full run we drop the initial logs.
Change-Id: Id2626f9113d82c6d524039acda8a8ec74afb2081
If you have
devstack_localrc:
ARGUMENT: "argument with spaces"
The quotes get lost during YAML processing and the resulting file has
ARGUMENT=argument with spaces
which is a shell error.
Quote all arguments to avoid this sort of thing.
Change-Id: Ia63a53d745dfea7262bcdb5d46425f431c3ccfe5
OpenSSL 1.0.2 generates key files with default permissions: 644 and the
files are copied to the /etc/pki/* directories with sudo.
When the default CI node Ubuntu version was changed from Xenial =>
Bionic we changed from OpenSSL 1.0.2 => 1.1.0. And OpenSSL 1.1.0
generates key files with default permissions: 600. When we copy the key
file to /etc/pki/* using sudo, it becomes owned by root and then the
console-related users are unable to read it.
This sets the ownership of the /etc/pki/<console> files to the
user:group intended to read them.
Closes-Bug: #1819794
Change-Id: I437a46c875cf633272e8cad0811e5557f2ac3641
I7d16194d6ba1391ca31251d5b50cbb8de033fc38 added wrong behavour
on Fedora > 26 and Centos 7 when python3 disabled
pip should install packages in /usr/bin
Closes-Bug: #1820070
Change-Id: I3a8efbc8eb6e311db9c7347577c5d2047ba523a9
This makes the grep match in check_python3_support_for_package_local
the same as check_python3_support_for_package_remote.
Change I0349de2026c49279ba7f262d5e86d37018d66326 in grenade started
setting the PYTHON3_VERSION variable, and then we recently started
using bionic nodes everywhere which means we're running python 3.6.
The etcd3gw package has a python 3 and 3.5 classifier, but not 3.6:
https://pypi.org/project/etcd3gw/
The pip_install function code that is dealing with installing py3
packages is hitting a problem installing etcd3gw if the package is
local because of the more restrictive grep in the
check_python3_support_for_package_local function, and since
PYTHON3_VERSION=3.6 now, we don't install from py3 and install
etcd3gw on python 2.7 which makes services like cinder-volume and
cinder-backup, which use etcd3gw, fail when they are running under
python 3 (they get module import errors).
This simply removes the $ restriction on the grep. Looking at the
change that added those local/remote functions:
I243ea4b76f0d5ef57a03b5b0798a05468ee6de9b
There is no explanation for the difference, it just said:
Also, since not many packages are classified correctly, fallback
to looking for just "Programming Language :: Python :: 3" and
log a message for the package to highlight the problem.
So that's what this change does.
Note that alternatives would be:
1. Update the etcd3gw package to add the 3.6 classifier and do
a release (this should probably happen anyway).
2. Add etcd3gw to ENABLED_PYTHON3_PACKAGES but that would be a
short-term hack workaround.
Change-Id: Icd3768870ba0f1659bb2e6f002043d975047b73e
Closes-Bug: #1820892
When running stack.sh locally on stable branches
with tempest enabled and TEMPEST_PLUGINS set,
devstack will try to fetch master branch of requirements
and that fails if branch is not tracked.
Change-Id: Ia1ae6869a8fede2af5cd7c875e0946b6a75eb518
Closes-Bug: #1820051
TEMPEST_PLUGINS contains the list of the tempest plugins installed
alongside tempest by lib/tempest.
If TEMPEST_PLUGINS is not explicitly set, the new tempest_plugins
variable is used to fill it by combining its items with
the base devstack path.
Change-Id: I9f1fa2755e16871ff9d6ba33fdeaf3023eedf8d4
Plugins must be the last items in the local.conf file
otherwise the configuration set in the rest of the file
is not applied to them (for example a different value of DEST.)
Change-Id: Ia001badca179c3f3436d5ecd26b0755a3f3a3078
The default version is 3.2.17 which seems to be too old.
Some external tools are not compatible with this old version.
For example, kubeadm cannot support external etcd version that is
older than 3.2.18.
This commit update the etcd version to 3.3.12 wich is the current
latest version.
Change-Id: Icfabbe580bb83a3babb98cc9fdbfb8eb388dc108
On CentOS/ Fedora machines, this can be useful when QEMU silently fails
to start up due to SELinux denials. For Debian-based machines, which
use AppAromor, DevStack already captures the output of 'kern.log' (via
`journalctl -t kernel` redirected into 'syslog.txt.gz').
Change-Id: I231b22664f0944b905e00568759785615a1d47c3
Acked-by: Clark Bolyan <clark.boylan@gmail.com>
Signed-off-by: Kashyap Chamarthy <kchamart@redhat.com>
This updates various parts of documentation to use https, rather than
git, which is not implemented by gitea
Change-Id: I8d2a93128dcdaba0a00b43d18652781733f90cf0
Story: #2004627
Task: #29701
Infra are looking at implementing gitea for serving git, but this does
not have a git protocol handler ATM. Switch GIT_BASE, and some
testing, to https:// to be in a better position to handle this.
Change-Id: I97a7b0de7b1ec2dd15d15c58699a631b09273df1
Story: #2004627
Task: #29701
The bridge-utils package has been deprecated for some time now [1] and
'brctl' does not exist on some more recent distros like Fedora 28.
Replace references to brctl with the proper ip commands.
Calls to "brctl show" are not being replaced with calls to "bridge link"
because the output format is very different and in testing some bridges
were not listed. So the simpler method of consulting /sys/class/net is
used.
In worlddump.py we try running both because failures are handled
gracefully by _dump_cmd(), as well as "ip link show type bridge" for
additional info.
[1] https://lwn.net/Articles/703776/ for example
Change-Id: Ie4c8ad6ce4a09c38023c9e4ec7834c249403145f
Partial-Bug: #1801919
Devstack has some little known variables for running nova
with the fake compute driver and running several nova-compute
services on a single host, which can be useful for testing
move operations and scale testing of controller services like
nova-api and nova-scheduler.
This adds documentation about the fake virt driver and related
variables and scaling considerations when using them.
Change-Id: Ic89d463d0f3f180b323edd6e2c8ff0404638ef07
when installing with python 3.6 on centos7 pip installs
packages to /usr/local/bin as it does on new versions
of fedora.
this change updates the check to include centos
Change-Id: I7d16194d6ba1391ca31251d5b50cbb8de033fc38
This introduces a breaking change in the URLs used to access the console
[1]. This is updated in both the documentation and linked nova change.
[1] https://github.com/novnc/noVNC/commit/83391ffc
Change-Id: I14a0be0034f4a76ab37eb90325967500c3bf1ff9
Depends-On: I9a50a111ff4911f4364a1b24d646095c72af3d2c
Related-bug: #1682020
Keystone is currently working through a bunch of changes to add proper
system, domain, and project scope support for its API. This includes
implementing ``admin``, ``member``, and ``reader`` roles for system,
domain, and project assignments. More informaiton on those specific
changes can be found here:
https://review.openstack.org/#/q/(status:open+OR+status:closed)+project:openstack/keystone+branch:master+topic:implement-default-roles
One thing that was uncovered in implementing that support for the
project API was that setting tempest
``CONF.identity.admin_domain_scope = True`` meant domain admins of one
domain would be able to list projects in other domains, highlighted in
the following patch:
https://review.openstack.org/#/c/624218/2
This commit doesn't set this option and assumes the proper
domain-scoping behavior being built into keystone natively.
Change-Id: I12a57cc43de0b17eababa19b7b94de5277689f82
Related-Bug: 1750660
when tempst venv is build, it use the master upper_contraint[1]
but when we install tempest plugin, it use branch upper_contraint.
This leads to mismatch the dependency version between tempest and required
tempest plugins setup.
Current flow after this change is:
1. install tempest form master (until you explicitly change TEMPEST_BRANCH
which is default to master in all stable branch). It applies the upper_constraint
from the stable branch but that will be overridden in step2
2. configure tempest, here the created venv will install all dependency with
master's upper_constraint.
3. install tempest plugins in same venv created above. Now tempest plugin
will also use the master upper_constraint.
With this tempest venv which has all enabled plugin will be contsraint with
master.
[1] https://github.com/openstack-dev/devstack/blob/72f632222f6d90d3545b5d7ca48297da4218e2ea/lib/tempest#L590
Change-Id: I89314e8391e8f26c622fc090cbe27997b3cf049a
Closes-Bug: #1816022
The read_password function is defined inside stack.sh
and it cannot be used inside the "public library interface"
provided by DevStack.
Move the calls found inside library files to stack.sh,
following the same pattern of the other calls to read_password.
Change-Id: I8adc6723b677dfac2bef735f660e056c498bf773
numpy is a python requirement of the websockify package (it appears
there was some disucssion over *removing* this in [1], but did not
happen). Possibly these packages were installed a long time ago
before wheel support as it was taking a long time to build. But we
have wheels today, and later versions than the distro provides are
being dragged in anyway. Remove all distro installs.
[1] https://github.com/novnc/websockify/pull/163
Change-Id: I322dd9e1a07d8ce03c26cf3fcccebd6e21282fe4
It seems nova has changed defaults on who can create zero-sized disk
instances [1] and now some devstack jobs, like nodepool's, can't
create cirros images using this flavor. It seems the easiest thing to
do is just to bump it up.
[1] https://review.openstack.org/#/c/603910/
Change-Id: I1172d4775d608568ccbeb27e2975d83add892ea9
Cinder v1 was removed over a year ago. Change the cinder template
URLs devstack defines in the glance-api.conf to use cinder v3
instead.
Change-Id: I4a68dc0b53631be0708e7411c37619dd6dfd4fa6
The default for VOLUME_BACKING_FILE_SIZE changes over time
and the docs referencing it are clearly not keeping pace so
rather than hard-code a default in the docs just remove it
since the doc already mentions the variable used to set that
size.
Change-Id: I4242584d13250872250689863d1b70c68594eefe
Cinder and etcd are enabled by default and by default
cinder uses etcd as a distributed lock manager with
tooz as an intermediary. We see a lot of ToozConnectionErrors [1]
in the cinder logs when etcd is backed up [2] which results in
cinder operations timing out causing test failures, like
when a volume is not deleted within a given time.
This changes ETCD_USE_RAMDISK=True by default to try and
alleviate some of the pressure. An alternative is if we know
we're in a single-node job we could just not use a DLM for
Cinder.
[1] http://status.openstack.org/elastic-recheck/#1810526
[2] etcd[26824]: sync duration of 12.076762123s, expected less than 1s
Change-Id: I5f82aa40e9d84114e7b7b5cf19ec4942d6552490
Partial-Bug: #1810526
Uses a less offensive and arguably better understood section
header in the multinode docs.
Change-Id: Ie6fd58e9abd5c1ce88d88ac55419807790f61851
Closes-Bug: #1810317
On SUSE Linux Enterprise distributions, lsb_release -i typically
returns "SUSE" not "SUSE LINUX" as the vendor string.
To avoid duplication of the same regular expressions in multiple
places, add is_opensuse() and is_sle() helper functions, and modify
is_suse to invoke those.
This may also be helpful in the future for distinguishing some corner
cases where things are handled differently between openSUSE and SLE.
Change-Id: I43bf163bc963758ddbb6289928837f5f6512f265
As the user on the node under test may not exist on
the zuul executor node we do not copy the log owner
or group to avoid the rsync task failing when it
tries to chown the files.
Change-Id: I500cf3692a4d27b0c2a0a4f5586580d180a8778e
The glance-api service may use multiple config files, so
tell oslo.config about the config dir instead of a specific
config file when the service is started.
Change-Id: Iad3602d209cbb31e10683c67e1fd6b465d19f560
Partial-bug: #1805765
The version comparison introduced in
I5152f2585c3d4d18853988d6290039d6b1713b99 was broken, because it tried
to use bash's -lt operator for floating point comparison, but bash
only supports integer arithmetic.
So instead use devstack's vercmp() function.
Change-Id: I8aac71c5bb6c2e82479d62831ea0672ba6a9a534
Older mariadb packages on SLES 12 provided mysql.service. The newer
ones on SLES 12 and 15 use mariadb.service; they also provide a
mysql.service symlink for backwards-compatibility, but let's not rely
on that.
Change-Id: Ife6bd007ba30af0b77d44832b19d518034bdb12b
The existing devstack guide for load balancing is out of date.
This patch updates the guide to reflect the current way to install
devstack with the Octavia plugin(s).
Change-Id: Id48b70b50e44ec7b965d969b2d93f77543d7364c
Seems like for etcd-heavy services like Kubernetes, the fsync
performance of gate VM's are too low [1]. This commit implements an
option to put etcd data directory on RAM disk (tmpfs) to work this
around.
[1] http://lists.openstack.org/pipermail/openstack-discuss/2019-January/001849.html
Change-Id: I5a17099cb9d6941b1a009dc82daefd2c7946d892
Recently iscsid was disabled by default on Ubuntu 18.04 (bionic),
and it may be on Xenial too, see:
https://bugs.launchpad.net/ubuntu/+source/open-iscsi/+bug/1755858
On a local Bionic deployment with Python 3, the lack of iscsid makes
nova-compute fail with an exception when trying to attach a volume:
Invalid input received: Connector doesn't have required information: initiator
Asking for the service to be started even if it is already running should not
hurt, so remove the check for the distribution.
This does not seem to be an issue on CentOS 7 (but Python 2) where
the socket activation of iscsid seems to work, so maybe there is
another way to make this working. Also, the service could be
enabled, not just started.
Change-Id: Ifa995dcf8eb930e959f54e96af6f5fce3eac28ae
Change I4820abe57a023050dd8d067c77e26028801ff288 removed access
to the database for the nova-compute process but only in
superconductor mode. Grenade runs in singleconductor mode though
so we are getting tracebacks in nova-compute logs during grenade
runs because nova-compute is running with nova.conf which is
configured with access to the nova API database.
This change handles removing database access for nova-compute
generically to cover both the singleconductor and superconductor
cases.
Change-Id: I81301eeecc7669a169deeb1e2c5d298a595aab94
Closes-Bug: #1812398
This allows plugins to specify their binary dependencies in bindep
format.
Some thinking on the implementation: this is in contrast to the
files/[deb|rpm] installation, which is called from the external
install_prereqs.sh script. This script being an externally callable
entry-point is really an artifact of the days when we would build
snapshot images for CI and wanted to pre-cache downloads. These days
we use the mirror system to keep packages close to CI nodes. Thus
rather than expand install_prereqs.sh to also be installing
virtualenvs and python dependencies, this seems to fit better as a
separate internal phase of stack.sh.
Documentation is updated
Change-Id: Icbdfbf97c17c906a7ae86f43e80eb2c445816228
This change addresses a few inconsistencies in how nova processes
are configured to speak to the placement service.
The initial inspiration was that region_name was not being set in the
[placement] section, despite $REGION_NAME being used when setting
the endpoint in the catalog. That's fixed.
While fixing that two other issues became clear:
* Configuring nova process to use placement should happen in lib/nova
not lib/placement so the function has been moved.
* auth_strategy is not relevant in the [placement] section of a
nova process
The name of the function is maintained, in case there are plugins which
call it, but a comment is added to indicate that other services besides
nova compute (such as the cell conductor) may use the function.
Change-Id: I4a46b6460596e9a445bd90de2d52dbb71fb963df
This adds a -bindep option to the key development library install
functions. With this option the bindep.txt file will be referenced
and the relevant packages installed.
Change-Id: I856f1f59fca49b6020920d8f859b797f3b904300
Given the file to be configured, if user "stack" even doesn't have
read access, the result of configuration is not expected. iniset with
"-sudo" option will always create the section and the option which we
want to configure for each calling, no matter whether this section and
this option exist in the file or not. The root cause is the calling of
grep and ini_has_option in iniset don't use the "sudo" option.
Change-Id: I9d21322046b7be411c4c7c28fefc24894fa2e131
Signed-off-by: Yi Wang <yi.c.wang@intel.com>
Remove the requirement that services explicitly enable python3 support
in order to be tested under python3 when running with python3
enabled. Keep the enable_python3_package() function for backwards
compatibility, for now, since it is called in some devstack plugins.
Explicitly add swift to the set of packages that should not be installed
using python3 by default until full support is available.
Change-Id: I8ab0a7c242bbf5bf3f091f5a85a98e2f4543f856
Signed-off-by: Doug Hellmann <doug@doughellmann.com>
There are already devstack plugins that contain a hyphen in the name,
like `networking-baremetal`. In order to allow ordering for these to
work properly, amend the regexes we are using to match any
non-whitespace characters instead of only alphanumerics.
Amend the test to cover this use case.
Change-Id: I91093a424f8d5e8007f140083e1ea36a81fe849f
Closes-Bug: 1809016
The test_plugin_deps function in the test code for the
write-devstack-local-conf role was missing the import part of actually
executing the code under test and asserting the expected result.
Change-Id: I125870b13d2581cdec0dede11157b19b702565cd
Keystone is moving more things to require a system scoped token to
work. Getting one of those requires that domain and project information
are not set.
Change-Id: I2e1640e9f9ef6cdf56bef49d1ae8f0591570c3e6
This commit fixes a grammar issue in the LDAP integration guide
and it adds prompts to the command-line examples to be more
explicit about where or how commands are being run.
Change-Id: Ic6a5adfbcf2841656929e6c3875889a31d314089
'integrated-gate-py35' template is going to be
renamed to 'integrated-gate-py3' in https://review.openstack.org/#/c/626078/
Integrated jobs are running on Bionic now where python 3.6 is available.
Which means gate jobs in 'integrated-gate-py35' template are
running on python 3.6 not on 3.5 which makes this template name confusing.
depends on commit rename the 'integrated-gate-py35' to 'integrated-gate-py3'
so that it can convey that template will use available python 3 version
in used distro. For example: 3.5 in xenial and 3.6 in bionic and so on.
This commit starts using the new template name so that old
template name can be removed.
Depends-On: https://review.openstack.org/#/c/626078/
Change-Id: I07048817eb826337dd5bd89a97711bb9d43495cf
gitignore is not parsing regex, only shell globs,
so '^' has no meaning, and local.conf is being thus tracked.
This patch properly ignores only local.conf in root of repo but still
tracks samples/local.conf and others.
Change-Id: I93ef778f1f3ee8101ce21cce377f7b527b7153f3
This enables direct-io on the loop devices that we create for LVM backing
stores. The goal here is to reduce the buffer cache overhead involved with
loop mounting a very large file on a filesystem, as well as potentially
providing a little more block-device-like behavior for things that expect
them. We are hoping this will address some of the very long LVM calls that
cinder does, which randomly take a very long time, causing timeouts.
The loop direct-io support was added in kernel 4.4.0, which was xenial,
but the losetup binary does not have the required flag. Thus, this patch
checks the "losetup -h" output for the flag before deciding to enable it.
Change-Id: Idc69cf3598d6ed6646c0145733c90ad0b1b60883
Volume API v2 has been deprecated for a long time.
There is no reason to use volume v2 in clouds.yaml by default.
This commit also drops "--os-identity-api-version 3" from
write_clouds_yaml in functions -common as "3" is the default value
of tools/update_clouds_yaml.py. They are hardcoded in DevStack
so there is no reason to pass it.
Change-Id: Ie84026a3d19f7711fc781b7012355096c7ff6b5a
This does a few things to the home page and all-in-one single
machine install guide:
* Uses code blocks for formatting
* Adds the customary "$" to the console blocks in the
all-in-one single machine install guide
* Instructs to use "sudo su stack" and adds a note about
"sudo visudo" in the all-in-one single machine doc
* Creates a symbolic link to the sample local.conf and links to
it from the install guide (note that local.conf might be old
by now)
* Fixes the .gitignore file to only ignore local.conf in the root
of the repository, otherwise it would ignore local.conf everywhere
including the samples and doc/source/assets directories.
Change-Id: I50ae7bd32c4c1caa2ac8551fc54b31dd2dfae568
This is the next release in the 0.3.x stable series, containing a fix
for getting out of disk errors when cirros reads metadata from a
config-drive[0].
[0] https://bugs.launchpad.net/cirros/+bug/1808119
Change-Id: Id2f20ebafdd78c2dadf81b8f80f22e7bd6db7755
We've run into what appears to be a race with apache trying to reuse a
pooled connection to a backend when that pool connection is closing.
This leads to errors like:
[Fri Dec 07 21:44:10.752362 2018] [proxy_http:error] [pid 19073:tid 139654393218816] (20014)Internal error (specific information not available): [client 104.130.127.213:45408] AH01102: error reading status line from remote server 127.0.0.1:60999
[Fri Dec 07 21:44:10.752405 2018] [proxy:error] [pid 19073:tid 139654393218816] [client 104.130.127.213:45408] AH00898: Error reading from remote server returned by /image/v2/images/ec31a4fd-e22b-4e97-8c6c-1ef330823fc1/file
According to the internets this can be addressed (at the cost of some
performance) by setting the proxy-initial-not-pooled env var for mod
proxy. From the mod_proxy docs:
If this variable is set, no pooled connection will be reused if the client
request is the initial request on the frontend connection. This avoids the
"proxy: error reading status line from remote server" error message caused
by the race condition that the backend server closed the pooled connection
after the connection check by the proxy and before data sent by the proxy
reached the backend. It has to be kept in mind that setting this variable
downgrades performance, especially with HTTP/1.0 clients.
Closes-Bug: #1807518
Change-Id: I374deddefaa033de858b7bc15f893bf731ad7ff2
Nova is moving nova-cells-v1 to its experimental
queue set of jobs so the comment in devstack should
be updated.
Depends-On: https://review.openstack.org/623538
Change-Id: Iefbaa9b809d1426640cbd47a42213f28c9ec5ff3
Related-Bug: #1807407
Enable the Software Collections (SCL) repository for CentOS. This
repository includes useful software (e.g. the Go Toolset) which is not
present in the main repository.
For example, Octavia uses a Go based testing tool and its CentOS-based
jobs got broken now with the update to CentOS 7.6 which no longer
provides golang.
Change-Id: Ic68a6d6cd7da41510e624b6bea7976d9a960af98
... even when no other subject alt names provided
Previously, a non-voting job in barbican's gate would fail with something like
X509 V3 routines:X509V3_parse_list:invalid null name:v3_utl.c:319:
X509 V3 routines:DO_EXT_NCONF:invalid extension string:v3_conf.c:140:name=subjectAltName,section=DNS:pykmip-server,,IP:198.72.124.103
X509 V3 routines:X509V3_EXT_nconf:error in extension:v3_conf.c:95:name=subjectAltName, value=DNS:pykmip-server,,IP:198.72.124.103
because we'd have an invalid empty string.
Change-Id: I5459b8976539924cd6cc6c1e681b6753a76b804c
Current code assumes the variable is being set to either "True" or
"False", which will lead to weird errors if it is being set to something
like "true" instead.
Change-Id: I88983c9150efad882cd867c2d14d86ba6b2522c9
Switch the nodesets that devstack job run on from Xenial to Bionic,
i.e. the latest Ubuntu LTS release. Keep variants running on Xenial
in order to make sure that we stay backwards compatible while we keep
running Xenial jobs on the older stable branches.
Change-Id: I8749ed24d5f451d29f767ebb2761abd743b7d306
Allow other jobs to explicitly require a node running Xenial. This seems
clearer than having a generic openstack-single-node nodeset which
implicitly uses Xenial.
Change-Id: I013fb8abd4e6ab6539bd9410acbc8446e57ec70c
Earlier review [1] suggested some cleanups which have been
done here:
* Removing a redundant call from cleanup_placement
* Fixing a typo in a comment
[1] https://review.openstack.org/#/c/600162/15/lib/placement
Change-Id: I8abd2f02b123c6c1937c026ff13eb4e600de3202
We introduce and set PLACEMENT_REPO, add a placement-manage command
to sync database tables (see one of the commits on which this
depends), use /etc/placement/placement.conf for config, and put the
uwsgi config file (pointing to placement-api instead of
nova-placement-api) in /etc/placement.
openstack/placement is also added to the required-projects in
the devstack zuul job.
Change-Id: I0b217e7a8c68a637b7a3445f6c44b7574117e320
Nova change https://review.openstack.org/603910/ is
going to change the default rule on policy
os_compute_api:servers:create:zero_disk_flavor to
admin-only, which will prevent non-admins from
creating image-backed servers with a flavor that
has disk=0 since it's a potential security exposure.
Therefore we need the test flavors that are created
for tempest to use non-0 disk values. Since the flavor_ref
and flavor_ref_alt can be aligned to the image_ref and
image_ref_alt in tempest.conf, we get the image sizes
from glance (in bytes) and convert those to GiB disk
sizes for each flavor, respectively. Since we're using
Cirros images by default, we need to make sure to round
up otherwise we'd still have a 0-disk flavor.
There are lots of ways the math could be done here
using numfmt, bash, awk, bc, etc, but it's simplest to
write and probably easiest to read by using python for
the size conversion code.
Change-Id: I537c299b0cd400982189f35b31df74755422737e
Related-Bug: #1739646
Neutron is in a process to migrate to policy-in-code.
DevStack needs to be able to handle both cases with and
without policy.json in the neutron repo.
Note that nova assumes neutron API access with admin
so user_name:neutron needs to be included in context_is_admin
to make DevStack work properly. Hopefully this can be cleanup
but this is a separate topic from policy-in-code.
Needed-By: https://review.openstack.org/#/c/585037/
Change-Id: Id1b0600d92e839ade1790a15c372e82e8e16ee9f
The universe repository is not enabled when installing Ubuntu from an
ISO (at least for Bionic). This leads to some errors during the devstack
run that are not seen when running based on a cloud image which has that
repo enabled by default. Enable that repository unconditionally, the
operation is idempotent.
Change-Id: Ifcb7ecd78fb25ca2136f5848c19b74500e520873
Closes-Bug: 1792936
This reverts commit faaf96bfb1.
Ironic jobs were still using this option, it needs to be
switched to an alternative first.
Change-Id: I1683d7cfa81f5fe2497cc7045e87f8b20fed4968
In order to make sure not possible to introduce a change in tempest
which breaks the shared network compatibility.
Depends-On: I6e3e53c4ac26b4fef09fefb9c590dfa91f577565
Change-Id: Ib2e7096175c991acf35de04e840ac188752d3c17
Builds of opensuse-tumbleweed nodes are currently failing, so these jobs
are receiving NODE_FAILURE.
Change-Id: I3c2d73a150df009e7dadc76277be36eb72e0dfa7
Docs say that you require Fedora 24/25 to run Devstack, but Devstack
is working in newer versions. Update document to say that Fedora 28
can be used instead.
Closes-Bug: #1797239
Change-Id: Ie5227db9943e5ddb93cd37440165eabbae22f4fc
Signed-off-by: Laura Sofia Enriquez <lsofia.enriquez@gmail.com>
The legacy job legacy-tempest-dsvm-neutron-pg-full is now named
tempest-pg-full - using the new tempest and Zuul v3 frameworks.
Change experimental job to use new job.
Change-Id: If16397724fb4facd2a0db8148bdf7ba427ca10b6
Depends-On: https://review.openstack.org/609530
Apparently we're inheriting some database config from the main file,
which should not be set for nova-compute. If we're properly in superconductor
mode where we have a dedicated config for compute, remove those lines
if present.
Closes-Bug: #1797413
Change-Id: I4820abe57a023050dd8d067c77e26028801ff288
We can see that there is more demand on using ipv6 as the underlay
infrastructure to deploy new services, and OpenStack should be ready for
that.
These devstack ipv6 jobs are based on the work started by Jens Harbott in
https://review.openstack.org/#/c/608168/
Change-Id: I55bd067487665e5026e82a0737cb0f38a69499fb
Signed-off-by: aojeagarcia <aojeagarcia@suse.com>
Tempest have removed the volume-feature-enabled.api_v1
config options[1] and modified the default value of
volume-feature-enabled.api_v3 to True.
These config options not needed to be set from devstack
side.
[1] https://review.openstack.org/#/c/573135/
Change-Id: Ic35cf4482ab4d3c2e69348ec92568e68f6ea74ee
Leap 15.0 has been released May 25th, 2018 (see
https://en.opensuse.org/Portal:15.0 ) and we'd like to
transition devstack against it and remove Leap 42.3 from
the testing matrix. Leap 15.0 is newer than Leap 42.3 as
the numbering schema of openSUSE was changed.
Co-Authored-By: Antonio Ojea <itsuugo@gmail.com>
Change-Id: I078f9a2580160c564c33e575008516f5e92239d6
Dnsmasq and haproxy are used frequently by neutron and nova, apparmor
profiles can block some operations and the deployed cloud can't
work properly so some tests are going to fail.
Some openSUSE distros has apparmor enabled by default so we need to
disable it.
Change-Id: I30fda684effb09810643e58bf0b31a73d7d9b378
Signed-off-by: aojeagarcia <aojeagarcia@suse.com>
As a follow-on to I28aebffce6c5561360a9e44c1abc44b709054c30; make sure
we quote the error messages on the way through so they retain their
newlines.
Change-Id: I493317948264941b4788b100a0b0bc13d2698acf
Python2 match routines for x509 fields are broken and have to use
the DNS field for ip addresses.
The problem is that if you use ipv6 addresses in the DNS field,
urllib3 fails when trying to encode it.
Since python3 match routines for x509 fields are correct, this patch
disables the hack for python3, encoding the ip address in the
corresponding field only of the certificate.
Partial-Bug: #1794929
Depends-On: https://review.openstack.org/#/c/608468
Change-Id: I7b9cb15ccfa181648afb12be51ee48bed14f9156
Signed-off-by: aojeagarcia <aojeagarcia@suse.com>
Old-style test fails on Ubuntu when python3 enabled, with:
.../devstack/inc/python: line 52: [: 16.04: integer expression expected
Use bash-style test, which doesn't attempt to evaluate the RHS if the
LHS evaluates to false
Change-Id: If18031ab98c9060e5825c3a8d3c647bd3705cd9c
Closes-Bug: #1796174
It turns out that a host can have multiple valid default gateways,
something that's not common in ipv4.
This patches add supports for multiple default gateways in ipv6
environments.
Closes-Bug: #1786259
Change-Id: I30bf655f7160dd19c427ee79acdf145671a3e520
Signed-off-by: aojeagarcia <aojeagarcia@suse.com>
Legacy backup service support was recently dropped from cinder in
change I3ada2dee1857074746b1893b82dd5f6641c6e579 and we need to
adjust how we set the config option in devstack accordingly. This
updates the backup_driver option to specify a full class name instead
of only the module name.
Closes-Bug: #1794859
Change-Id: I3a72f38b564b8b83b233fccba7685833b6394d45
I'm switching tempest/dsvm jobs to run on
Ubuntu 18.04 LTS (Bionic Beaver) on openstack/manila,
and I believe this nodeset can be here
so other projects can use it too.
Change-Id: Ib8279cde3e14d5378f27254188ee14dbb0800428
Needed-By: https://review.openstack.org/#/c/604929/
The external_network_bridge option is deprecated/legacy and being
removed from neutron (see I07474713206c218710544ad98c08caaa37dbf53a).
This patch removes the external_network_bridge option iniset from
devstack scripts.
Change-Id: I4d9641cc9bb83719c9af1edabb89a63c4c2b1d96
That change introduces correct way of generating msg
for die in common systemd pitfalls.
Co-Authored-By: Szymon Datko <szymon.datko@corp.ovh.com>
Co-Authored-By: Piotr Bielak <piotr.bielak@corp.ovh.com>
Change-Id: I28aebffce6c5561360a9e44c1abc44b709054c30
This patch switches the CINDER_ISCSI_HELPER from tgtadm
to lioadm in openSUSE distros, as it increase the performance
and reduce the flakiness on some tests.
Change-Id: Ic3ee9c6baabe20f8f4d14246f6e29808796a5db9
Signed-off-by: aojeagarcia <aojeagarcia@suse.com>
Allow the setup-devstack-source-dirs role to accept a target role
to be setup - when available - for the repos.
Change-Id: Iebcba0d4be6d9d71b783e10a82c35a406afbd6bf
legacy-periodic-tempest-dsvm-oslo-latest-full-master
runs only on master, remove it. This needs to stay in project-config.
Change-Id: I81e66ddb0976bb4bb7a7cd8efbbae3bda551191d
the swift and devstack-plugin-ceph jobs have been renamed, follow rename
and use in-repo jobs.
Depends-On: https://review.openstack.org/543048
Change-Id: Idccc21e47b2cc04e5eeab4db7f7fb7cf156f8049
Given that Natty and other releases that don't use cgroups have been out
of support in Ubuntu for years now, it's high time we removed the
special case code block that sets up the cgroup mount.
Change-Id: I5403a4b1b64a95236b4dfcb66c35c594a3460cca
This is a mechanically generated patch to switch the documentation
jobs to use the new PTI versions of the jobs as part of the
python3-first goal.
See the python3-first goal document for details:
https://governance.openstack.org/tc/goals/stein/python3-first.html
Change-Id: I338fc71919a41ec890bcb5edd0552ec7eb680eb5
Story: #2002586
Task: #24327
This is a mechanically generated patch to complete step 1 of moving
the zuul job settings out of project-config and into each project
repository.
Because there will be a separate patch on each branch, the branch
specifiers for branch-specific jobs have been removed.
Because this patch is generated by a script, there may be some
cosmetic changes to the layout of the YAML file(s) as the contents are
normalized.
See the python3-first goal document for details:
https://governance.openstack.org/tc/goals/stein/python3-first.html
Change-Id: I9169d41d790ae874af29c8ceccf0c55ab0df7727
Story: #2002586
Task: #24327
During Rocky cycle, horizon updates the path of the wsgi wrapper
to the one recommended by Django [1]. The old path will be dropped
in the T release.
[1] https://review.openstack.org/#/c/561802/
Related-Bug: #1763204
Change-Id: Ie942518b587d193a7de55ffcc0a2848406146eb2
In the latest version of Contributing to
DevStack manual.There are a few words to
say that "tools/build_docs.sh is used to
generate the HTML versions of the DevStack
scripts".But build_docs.sh is not there
since Newton version.So I delete it for
good.
Change-Id: I69f7aa23e1efd8f8a63aa79628e67378d524e173
Before this change, only *.log and *.log.[0-9] patterns were
ignored, which was not enough. Examples of file names which were
not ignored:
devstack.log.2018-08-09-100547
wget-log
wget-log.1
Patterns *.log.* and *-log.* work for every log file generated by
devstack.
Change-Id: I6f0de5de74f196ab9df66cf3f2f969e53da01c22
Signed-off-by: Michal Rostecki <mrostecki@suse.de>
Set 'PUBLIC_INTERFACE' in local.conf, so the code will
be entered into _move_neutron_addresses_route of
neutron-legacy.
But if lack of sudo to run command arping, the information
"arping: socket: Operation not permitted" occurs. So add
'sudo' for 'ARP_CMD' of lib/neutron-legacy.
Change-Id: I8ac8a9bc2bbba049c45b28bf9b93d9a10e398fe6
Closes-Bug: #1783046
TEMPEST_AUTH_VERSION should be 'v3' or 'v2' not 'v2.0'.
To disable the identity v2 admin tests TEMPEST_AUTH_VERSION is
being compared with 'v2.0' which is incorrect.
Change-Id: I5f7e3bcf733edbbee06016bcad4845dda552815e
There is no projects using this and allows openstack-infra to delete
fedora-27 images.
Change-Id: I37d482dd2b5e099c370ab693ff430cb9c56360f8
Depends-On: https://review.openstack.org/588369
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
The RPC transport_url for keystone was being set in the DEFAULT
section, even though keystone doesn't do anything with it. Instead,
keystone leans on the [oslo_messaging_notification] section from
oslo.messaging to register the transport_url option.
This change sets the transport_url in the proper section instead of
using the DEFAULT section.
Change-Id: I11590d0175da7ea310d5529f2d7c0bf8d7fb25b3
This patch provides a new mechanism to deploy Neutron using
WSGI script. This also starts a Neutron RPC server process
when the Neutron API is loaded via a WSGI entry point to
serve the agents.
Co-Authored-By: Victor Morales <victor.morales@intel.com>
Co-Authored-By: Nguyen Phuong An <AnNP@vn.fujitsu.com>
Change-Id: I16a199b04858bfc03ef50d9883154dba8b0d66ea
Depends-On: https://review.openstack.org/#/c/580049/
Partially-implements: blueprint run-in-wsgi-server
The neutron guide refers to ENABLE_PROJECT_VLANS and
PROJECT_VLAN_RANGE but these are not present/checked in the code,
which uses ENABLE_TENANT_VLANS and TENANT_VLAN_RANGE. This
corrects the documentation to match.
Change-Id: I204356c861157e9fab357bb4dde55185bf18a707
The tempest-multinode-full job is running the c-bak
service on the subnode where swift isn't running, and
because of the "is_enabled_service swift" check, cinder
on the subnode wasn't getting configured to talk to
swift so the c-bak service was down. Since chances are
good that we're running swift, just configure cinder
to always use it.
Change-Id: I86b090967dadeeefc017ff0311beeea9441b6ba6
Closes-Bug: #1783128
Recently, Keystone renamed "Member" role to "member"
(case-sensitive) with https://review.openstack.org/#/c/572243/14
Case-sensitivity role requirement in Keystone was recently
formalized with https://review.openstack.org/#/c/576640/
From the above reference:
"Role names are case-insensitive. for example, when keystone
bootstraps default roles, it creates `admin`, `member`, and
`reader`. If another role `Member` (note the upper case 'M') is
created, keystone will return a `409` Conflict since it considers
the name "Member" == "member". Note that case is preserved in these
cases."
It follows that Tempest should use "member" role by default.
Change-Id: Iebf04fdb4c195b6779c74f66da3f7822cf174494
These seem to be not run for quite some time and they don't
succeed anymore - drop the code to avoid somebody accidentally
running it and wondering. A good example of "if it isn't tested
its broken".
Depends-On: https://review.openstack.org/583146
Depends-On: https://review.openstack.org/583147
Change-Id: I99e8a5ca2925217a5a2401984f3f4f6f032017be
token.provider.drvier.uuid and token.driver
has been removed from keystone[1].
Devstack has reference/setting of those config
options which is confusing for user and it can
lead to import error like[2]
This commit cleanup the devstack bits of removed
config options.
bp removed-as-of-rocky
[1] https://blueprints.launchpad.net/keystone/+spec/removed-as-of-rocky
[2] http://paste.openstack.org/show/725391/
Change-Id: I29b3b356622c485c4c1046679234a38e7b645071
When setting up a 3pci zuul, there is an edge case where a downstream
zuul may already have openstack/foo projects, eg:
review.rdoproject.org/openstack/foo. In this case, if openstack
projects are not namespaced to include the connection information zuul
gets confused and complains. We can avoid this by using the fqdn for
git.o.o for devstack jobs and both upstream and downstream zuul will
properly use the correct connection.
Change-Id: I01419ea9f51ce7491aa319b6240aec9c0d4f2356
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Don't specify hypervisor_type=vz property for hds images as nova
now reports only really connected hypervisors and not all supported
by a compute host
Change-Id: Ibec0856519ffe593d44d123f3b401eae19f7d95a
On some slow system with recent version of
Elasticsearch we saw that the service
becomes ready after more than 1 minute.
Change-Id: Id2b21ab24a96d10fffdcccd652a7d3ec4e8ce39c
With the move to flask, Keystone does not utilize paste-ini. This
patchset removes the paste-ini support from devstack for Keystone.
Change-Id: I8dd629937c9178660992fd648175dbef80ffa3c2
When we change the name of compute host then devstack is breaking
because it is using default host name from host.
How to change compute host name in local.conf
[[post-config|$NOVA_CONF]]
[DEFAULT]
host = foo
Change-Id: I4d4392f1f58f0431b10764610668565af88d392f
Signed-off-by: Prabhat Ranjan <pranjank@in.ibm.com>
The commit e95f2a3664 broke
networking-ovn (and potentially other ml2 drivers) by making the config
parameter mandatory. It doesn't need to be.
Change-Id: I0d5738ac3a6d27ddb7655835d77689409a6ff6f4
Bring along the required rpms for the ride on Fedora 28 (we really
should find a way to maybe do f* or something to avoid this
... consider it a todo :)
Change-Id: I37fd38de9baab478c86d23ea2cebca59dc8a5ed1
I'm not sure what the history of the (capital-D) Django package or
pyxattr; they've been there for a long time and should just be
installed like other dependencies these days.
Change-Id: I423230cc5cbb13d2cfb7b926a9571a8157ce5c46
The nova-conductor service running in the cell
needs to be configured to talk to neutron for
things like deallocating networks during server
build failure. This changes the configure_neutron_nova
flows such that the top-level nova.conf is configured
as before, but we also configure each nova_cell*.conf
cell conductor config files to also be able to talk
to neutron.
Change-Id: Ic5e17298996b5fb085272425bb3b68583247aa34
Closes-Bug: #1777505
Keystone now provides a set of default roles in addition to `admin`
by default [0]. This is done during the `keystone-manage bootstrap`
process.
This change aligns the `Member` role override from devstack with the
`member` role provided from keystone.
[0] https://review.openstack.org/#/c/572243/
Change-Id: I3da3530aa73a8a1500116bcefdcba7b947d5e05e
Closes-Bug: 1777359
This automatically always adds the project under test to LIBS_FROM_GIT
which effectively makes the normal "tempest full" job the same as the
"forward testing" job when it is applied to a library repo.
Change-Id: Ibbdd8a86e0ff55f67bef73e08e693b34a61b24df
This has all been around for a *long* time, like when dnf was a weird
new thing. Now it's the opposite and yum is a weird old thing :)
Choose it by default for platforms with it (Fedora, for now).
Change-Id: Id2bd7d145354b996de31944929fd0267ec24a08e
The function was using an undefined variable to show the version of
python3 being used.
Change-Id: Ibc956975d620ed5174de8823f9c202a680c56aaf
Signed-off-by: Doug Hellmann <doug@doughellmann.com>
We want to default to running all tox environments under python 3, so
set the basepython value in each environment.
We do not want to specify a minor version number, because we do not
want to have to update the file every time we upgrade python.
We do not want to set the override once in testenv, because that
breaks the more specific versions used in default environments like
py35 and py36.
Change-Id: Id83cb3cdd62517045c45388f88cb3de0e3d75da1
Ceph for example uses them. Creation already worked, but not
updates of existing keys.
Closes-Bug: 1774956
Change-Id: I20cb61c08079b9cd9ad56ac875525abf1442bff6
This makes three changes:
1. The quota options set when using the fake
virt driver have been renamed so we're getting
deprecation warnings on using the old names.
Rather than set each quota limit value individually,
we can just use the noop quota driver for the same
effect.
2. The enabled_filters list for the scheduler was last
updated when using the fake virt driver back in Juno
via Ic7ec87e4d497d9db58eec93f2b304fe9770a2bbc - with
the Placement service, we don't need the CoreFilter,
RamFilter or DiskFilter. Also, in general, we just
don't need to hard-code a list of scheduler filters
when using the fake virt driver. If one needs to set
their own scheduler filter list, they can do so using
the $FILTERS variable (or post-config for nova.conf).
3. The largeops job, which ran the Tempest scenario tests,
has been gone for a few years now, as have the Tempest
scenario tests, so the API_WORKERS modification when
using the fake virt driver should be removed. If we had
a CI job like the largeops job today, we would set the
worker config via the job rather than in devstack.
Change-Id: I8d2bb2af40b5db8a555482a0852b1604aec29f15
We use $API_WORKERS to throttle the number of workers
in other services but were not doing it for g-reg for
some reason, which by default will run ncpu workers
up to a limit of 8.
Change-Id: Idc81ce05546e6d625c10e2229256eafbe7c057a5
Closes-Bug: #1774781
Some commands must be run as a separate group to work. Users can use the
'sg' tool to do this.
This may be assumed knowledge for many users but it's helpful to note in
this, the definitive resource for DevStack's systemd integration.
Change-Id: I271c1d21b44fa972c152780c1caa01c21c265159
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
The same code is already in place for the new lib/neutron library, allow
this functionality to be used also when the neutron legacy services are
still being deployed. This was mangled in [0].
[0] I868afeb065d80d8ccd57630b90658e330ab94251
Change-Id: I7214c4893943fbfbeb42ad140f433eecd6c3e9f0
dosfstools provides mkfs.vfat which is needed if n-cpu is
configured with 'config_drive_format=vfat'.
Change-Id: If1e1537a079e71847d91ae03ed0c18290a467c4e
Related-Bug: #1770640
openvswitch firewall has been in Neutron tree since Newton and has gone
through lots of improvements since including simple upgrade path from
the iptables hybrid driver.
We have a tempest job running in Neutron tree with openvswitch firewall
that's been voting and stable for a while. For neutron_tempest_plugin,
we have had the openvswitch firewall in use since the beginning.
This patch proposes openvswitch firewall driver to become a default
driver for openvswitch agent deployments.
Change-Id: If26d0180e459210511f25f1faa83dd8ccea25ff4
Change 12579c3db7 moved console-related
settings from the global nova.conf to the per cell nova_cellN.conf
because of a recent change in nova that moved console token
authorizations from the nova-consoleauth service backend to the
database backend and thus changed the deployment layout requirements
from global console proxies to per cell console proxies.
The change erroneously also removed console configuration settings from
the nova-compute config file nova-cpu.conf because the nova-cpu.conf
begins as a copy of the global nova.conf.
This adds configuration of console proxies to the nova-cpu.conf in the
start_nova_compute routine. The settings have also been split up to
clarify which settings are used by the console proxy and which settings
are used by nova-compute.
Closes-Bug: #1770143
Change-Id: I2a98795674183e2c05c29e15a3a3bad1a22c0891
* v3 is a superset of v2 and has been the defacto Cinder version for
several years now.
* Devstack installs Cinder v3 API by default, so the default environment
variables should reflect this.
Change-Id: I86e1ae4e020e2be043cf8e190d7959b65b6c093c
Change 969239029d4a13956747e6e0b850d6c6ab4035f0 completed the
conversion of console token authorization storage from the
nova-consoleauth service to the database backend. With this change,
console proxies need to be configured on a per cell basis instead
of globally.
There was a devstack change 6645cf7a26
following it that re-enabled the novnc tempest tests, but the nova-next
job that runs the console proxies with TLS is *not* part of the normal
set of jobs that run on devstack changes (it's in the experimental
queue), so it was able to merge without the nova-next job passing.
This configures the nova console proxies in the per cell configuration
file if cells v2 is configured for multiple cells in order to pass the
nova-next job.
Closes-Bug: #1769286
Change-Id: Ic4fff4c59eda43dd1bc6e7b645b513b46b57c235
To help avoid the amount zuul.yaml chrun when we bring a new version
of fedora online, switch to using fedora-latest. As of writing,
fedora-28 is the latest release which we update our testing for.
Also add fedora-28 support to stash.sh and remove fedora-25 /
fedora-26 as they are EOL.
Change-Id: I3d716554e8f270f4434cc9cac3408f8e890e0665
Depends-On: https://review.openstack.org/565758/
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Nova has dropped support for non-resource class
baremetal scheduling, so the IRONIC_USE_RESOURCE_CLASSES
flag is no longer useful and has been removed.
Depends-On: https://review.openstack.org/565805/
Change-Id: Ib2e6c96409c98877f6a43b76f176c1420d2d415e
In a multinode setup, the compute node needs to report to the
placement service. If it does not do so, it effectively does not exist
from the scheduler's point of view. This patch adds placement-client
to the compute node's ENABLED_SERVICES so that this can happen.
Change-Id: Ibfcd84e4626301bcdea70f719ade7f8365d03497
cinder does not yet support operations without project_id in the url.
The unversioned endpoint is not a usable endpoint for a user that
requests the block-storage service. Although it would be lovely to have
the block-storage service have the unversioned endpoint in the catalog,
we need to get project-id out of the urls first.
Change-Id: I4246708b6ea31496ba4d565ab422abc76f730ee7
Needed-By: https://review.openstack.org/564494
Once the nova patch series that converts from the nova-consoleauth
backend -> cell database backend lands, we can re-enable the novnc
tests in tempest.
Depends-On: If1b6e5f20d2ea82d94f5f0550f13189fc9bc16c4
Change-Id: I2939191a1c3ce49fa2104b4ffdf795fc416a1c33
Along with converting to the database backend for console token auth,
the console proxies need to run per cell instead of globally. This way,
the instance UUID isn't needed in the access url as users will be
handed an access url local to the cell their instances is in. With
console proxies sharded across cells, a large cloud will no longer have
a bottleneck of one console proxy for the entire deployment.
This also disables the novnc tempest tests with a TODO to re-enable
them once the nova patch series that converts from the nova-consoleauth
backend -> cell database backend lands.
Change-Id: I67894a31b887a93de26f3d2d8a1fa84be5b9ea89
With change I7e1e89cd66397883453935dcf7172d977bf82e84 the placement
service may optionally use its own database. In order for this to
work, however, the ordering of how both nova and placement are
configured and initialized in stack.sh requires careful control.
* nova.conf must be created first
* then placement must make some adjustments to it
* then lib/placement needs to create the placement database
* before nova does a database sync (of both databases)
Otherwise, when the placement_database/connection is defined, the nova
db_sync command will fail because the placement database does not yet
exist. If we try to do a sync before the nova_api database is created
_that_ sync will fail.
This patch adjusts the ordering and also removes a comment that will
no longer be true when I7e1e89cd66397883453935dcf7172d977bf82e84 is
merged.
Change-Id: Id5b5911c04d198fe7b94c7d827afeb5cdf43a076
The n-novnc service only runs on the controller node, however novnc
settings must be enabled on both nodes for vnc to work, since both
hosts are compute hosts.
Change-Id: Icc29441f507e6e4df9fd900eb7f35b0862f52043
For compute migration to work, the stack user needs to be configured
with passwordless ssh between all hosts involved in the migration.
Reuse the build ssh-key for this, which is already distributed for
user root.
Depends-on: https://review.openstack.org/563584
Change-Id: Id07f55fea06509466add35315c135dbfba6aa714
This commit just makes sure that the configuration file for keystone
exists on the system. We use iniset to actually populate the values
we want before we run keystone anyway.
This results in a cleaner configuration file that isn't bloated with
comments and help text.
Change-Id: I7a1f879e9e242a11e2c4663ec116e33da28db7f5
If a project shows up in zuul's required-projects list, add it
to LIBS_FROM_GIT automatically. This way, when a user specifies
that a job requires a zuul-project, it gets used in testing, but
otherwise, it doesn't (pypi is used instead).
Also add information about what happens behind the scenes for both
LIBS_FROM_GIT and plugin dependencies.
This moves the check performed in check_libs_from_git to
a helper function which is installed for most kinds of
installations. This means that if someone sets LIBS_FROM_GIT to
"foobar", devstack won't error anymore, as nothing is going to
try to install foobar, therefore the check won't run on that.
However, as we move to automated generation of the local config,
that error is not likely to happen. This check was originally
added due to an error in the upper-constraints file (where a
constraint name did not match a package name). This location of
the check would still catch that type of error.
Change-Id: Ifcf3ad008cf42d3d4762cfb3b6c31c93cfeb40db
It looks pip 10 failed the uninstallation of distutils installed
packages. This patch temporarily cap the version of pip to work-around.
Closes-Bug: #1763966
Change-Id: I8bf80efc04883cd754c19bea0303064080112c6e
This commit applies the constraints for the tempest plugin installation
so they won't go over the upper reqs.
Closes-Bug: 1763436
Change-Id: I5cf91157bbdae79dec01d5b3db32efea21f1b2b7
In Tumbleweed genisoimage was dropped in favor of cdrtools,
so installing that no longer works. We can however install
mkisofs directly and switch to that as that is also available
in Leap 42.3 and Leap 15.0+ family distros.
Also drop dependency on libmysqlclient-devel which appears
unnecessary (and is no longer available with mariadb 10.2+)
Change-Id: Ie8402204b6cdf94c21865caba116d3fd1298c5ad
There is currently a OVS 2.9.0 update in Tumbleweed that
fails to start as it is having a race with systemd on creating
the home directory. Workaround is to run it as root for now.
Change-Id: Ief610c6473834b02a1d644d8f50d11138a48e6e6
In Queens and later, the application credentials feature is available on
keystone and enabled by default. It should be tested in devstack.
Depends-on: https://review.openstack.org/545627
Change-Id: I4b0dc823487e79df16e1e603012ba4a7dc438389
guides/devstack-with-lbaas-v2 contained an indentation misstake,
that formatted some of the text unintentionally as quotations.
Change-Id: Ibbad4974c45f028d3de461ba69e0cea837d9c871
The [placement]/os_region_name config option is deprecated
and no longer required to be set (the default is fine for
devstack) with the dependent nova change.
Depends-On: I973180d6a384b32838ab61d4e6aaf73c255fd116
Change-Id: I6379acf179ed511f1cdadbd7fb09e2454182a5d3
Devstack supports deploying an LDAP server and configuring keystone
to use it, but we didn't have any documentation for it. This commit
adds some basic documentation that should help developers setup
LDAP-backed development environments.
Change-Id: I8ba07d73f52cb7f575ff2953977e9fdcade92d83
For folks who are doing functional testing with less than the full set
of normal base services. Should be a no-op/ignorable for most people.
Change-Id: If14ee018c01995e0a5b6bcdaac9ddc8810c6d503
This makes sure that it is available to subprocesses like the other
authentication data.
Change-Id: I513b7c2620b171ce20a1ceb5536226f3a69f2b82
Closes-Bug: 1760901
Fix a few path issues where we didn't properly use NOVA_BIN_DIR /
SWIFT_BIN_DIR.
This is part of the effort to start using a virtualenv for openstack
services.
Change-Id: I6eb383db65cc902c67c43e5cb1a16a9716a914b2
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
In I5103b4331a8d7c5660848fc148ebe4139ce6dad9 it was noted that the
comment was wrong. While this has no functional impact let's clean it
up while it's fresh.
Change-Id: Ia6cf8125214c69f8289fa7cff948afc10801ed2f
This updates the default ETCD_VERSION to the latest 3.2 etcd release,
v3.2.17. 3.2 is chosen as it is packaged in bionic and fedora; we
hope to move to packaged versions for distros that support it in due
course.
This version supports arm64 and ppc64le which were not supported by the
previous default, v3.1.10.
We have removed the override to tarballs.o.o, as these files are now
cached as described in [1]
[1] http://lists.openstack.org/pipermail/openstack-infra/2018-March/005871.html
Depends-On: https://review.openstack.org/556688
Change-Id: I5103b4331a8d7c5660848fc148ebe4139ce6dad9
When nova-manage db sync runs on cell1 in superconductor
mode, the [api_database]/connection config option isn't
set in the config file on purpose so the cell can't
reach the API database.
As a result, the db sync on the cell config can't hit
the API DB to sync cell0, which is not something we need
here anyway, but it results in an error message.
This tells the cell config db sync to just run it on the
cell database and not try to sync cell0.
Change-Id: Iac092762decd6de9e90e264f2998d255e8e40d00
DEFAULT_VOLUME_GROUP_NAME volume group is LVM ephemeral storage used by
Nova. It is created by init_nova() if user sets NOVA_BACKEND to "LVM".
However, init_cinder() is also hardcoded to create it, based on the
asumption that CINDER_ENABLED_BACKENDS includes it. That assumption
doesn't hold for the current code. What's more important, even if user
wants to use DEFAULT_VOLUME_GROUP_NAME as one of cinder backends and
adds it to CINDER_ENABLED_BACKENDS, the current code in init_cinder()
are general enough and should work fine. This change removes relevant
code in init_cinder(). It also moves DEFAULT_VOLUME_GROUP_NAME clean-up
code from unstack.sh to cleanup_nova().
Change-Id: I53762f8eda6256f962cc4e1f1098406879bbcf5c
This updates the UCA usage from Pike to Queens. As a result,
the various volume multiattach checks can also be removed
because the Queens UCA has libvirt 4.0.0.
Change-Id: Icb971831c8d4fe5f940d9e7993d53f1c3765e30f
This PS adds the project_tags identity feature flag which allows
identity v3 project tags API functionality to be enabled for
releases after Pike. Once Pike is no longer supported in Tempest
this feature flag can be removed.
Depends-On: Ibaec1df79c9ac69c65cf5075c3519092bc609546
Change-Id: Iec6b34c10ea1bd7103720c773b48ce130643115d
I had to dig into the stack.sh code to figure out why
my tempest.conf post-config wasn't work, and it's because
post-config isn't the thing to use to configure tempest,
test-config is.
Change-Id: Ic5bbe36b5d44880d0a3a602f653b4f61fd89e9c8
Related-Bug: #1755947
Add the upcoming release Ubuntu 18.04 Bionic Beaver to the list of
supported distros. Drop the now unsupported 17.04 (zesty) instead.
Change-Id: Iea0b4bfdc510797f7886fac96eff6fdfb730252d
The patch to remove the use of the test-matrix [0] also switched from
using the neutron-legacy based service names (q-*) to the new neutron-*
names. However it turns out that the new implementation is not yet working
properly for most neutron-consuming projects, so we switch back to the
previous situation for now.
[0] https://review.openstack.org/546765
Change-Id: Id6de87211d6c4ea8fd14aa9203d8d5b17e9e2f04
Cinder supports both noauth and keystone auth mode. So now we can
configure this value via local.conf:
[[post-config|$CINDER_CONF]]
[DEFAULT]
auth_strategy = noauth
Change-Id: I1e434362117ab30dae71a8f3a80bc139e78f51bc
The -C option is not available in git versions older than "1.8.5" which
are still shipped by several distributions including centos 7.
Due to this incompatibility the patch has broken third party CI for
Cisco on Ironic.
Change-Id: I09a6f83f8b2fee870e6e1c50cbfdf2da4d70dfb2
This adds the necessary fixes to pass a devstack run
on openSUSE Tumbleweed. Also removes opensuse 42.2 as it
is EOL for some time already and no longer actively tested
in the OpenStack infra.
Depends-On: I1b68c08c07cf6653ea58506f738cbe0054b38f3a
Change-Id: I2894482deef063fd02b0818c695a2ddbf6767039
Document that orchestrate-devstack requires a linear strategy in the
invoking play. Also enforce the strategy in devstack.yaml.
Change-Id: Ia081225ec2be959fc5a4ddfd491f526296a8ca10
- There are some locations where we need the raw IPv6 address instead of the
url-quoted version enclosed in brackets.
- Make nova-api-metadata service listen on IPv6 when we need that.
- Use SERVICE_HOST instead of HOST_IP for TLS_IP.
Change-Id: Id074be38ee95754e88b7219de7d9beb06f796fad
Partial-Bug: 1656329
Document how to map DEVSTACK_GATE flags into the zuul v3 ansible
world. This is just an initial structure, the idea is to document
most of the flags as well as provide example in-line and links to
finished jobs.
Change-Id: I377ebb529bcd8f4971906563c577e8cfc48b98e6
Add inline documentation to jobs and render it in the jobs doc page.
Adjust the roles page to match jobs for title and filename.
Change-Id: I47a2b4b379c8517b0dea59a75943f3f871c29046
Extend the devstack job so that it can support both single and multinode
cases. Multinode mode require extra settings in devstack configuration,
some of which as subnode specific, some controller specific.
Also keep a simple devstack-multinode job defined for now so we can run
a multinode job in devstack gate, until the full tempest multinode job
is ready to match the old
gate-tempest-dsvm-neutron-multinode-full-ubuntu-xenial-nv.
Fixing multinode also requires sharing the CA configuration between
controller and peers, overlay network configuration for communication
between virtual machines and running discover_hosts for nova after the
subnode has been setup.
The extra orchestration required for multinode is encoded in a
dedicated role to allow for jobs in other repos to re-use it.
Change-Id: I2dcbd9bdb401860820e655d97aa3c4775af2827f
The openstack-ansible team found a regression in Queens when setting
the compute RPC upgrade_levels to 'auto' on a fresh install before any
computes had started up. The dependent change fixes the issue in nova
but for future proofing against this sort of issue again, we can set
the compute RPC upgrade levels in devstack to 'auto' for fresh installs
as well. Note that grenade already sets 'auto' for compute upgrade
levels, which is why we didn't catch this in grenade testing with the
compute RPC 5.0 version bump that caused the issue.
Depends-On: https://review.openstack.org/549737/
Change-Id: I07f34dbc09b6108ba8f5b2a83a28c75eb42be495
Related-Bug: #1753443
The legacy-tempest-dsvm-cells job is being moved into the
nova repo and renamed to nova-cells-v1. This change adds
the new job name to the in-tree definition of the experimental
queue jobs that run on devstack changes. A project-config change
will depend on this patch to undefine the legacy job name from
being used on devstack changes.
Depends-On: https://review.openstack.org/549780
Change-Id: I22fa1411809c46ffc423e0dd1cde0d8f40362635
With I4161e1f1c8d47070dd35fad38b00715438d94eb2 and
I37fe007dc6f387d43cbaf55771027718005ac40d we have removed most of the
legacy experimental jobs.
Add a short note so we don't go back to the Hotel California model of
"you can check in but never check out". With zuulv3 it is possible to
have self-testing changes for testing indiviual jobs. Experimental
should be for things with wider application to run against all
changes.
Change-Id: Ibfb902b17a8f7d5355689f2a584c061c001df0d8
The role that sets up the user and its home folder must ensure that
the home folder is owned by stack as well.
Change-Id: I2e72d7b9d68a2a14f8a148ef82cbb3f569bd1cea
Emit a disable_all_services and define which service we run in
the job directly. This drops the dependency from the test matrix,
from devstack default list of services and it makes it easier for
jobs to add/remove services based on the list in the base job.
Change-Id: Ib1debefd541b933dbfc54d484c263cc0ed60423d
/tmp is a world writeable directory, so using hardcoded filenames
in there is just a bad coding style (susceptible to symlink attacks).
Avoid using it to not give a bad precedent.
Change-Id: Ia66763a0e4714f2226e98dbd85600b2035bd5088
Cinder change I5231f8fe3399deb9c57e6efb121d0d008dc9c7f4
replaces iscsi_helper with more general one.
Change-Id: I49fe0365b170e5a5b0449d80003bcf970e4c191d
the shell script used is actually being run thru 'sh', not bash,
which does not understand "[[" test operators.
Explicitly run this script with /bin/bash instead.
Change-Id: I551d2631bcb6aef49550d69b3830ffcb509abfb7
Define an abstract job devstack base that does not require any
project apart from devstack. This job defines basic devstack_localrc
settings that are common to any devstack job (mostly to work with
infra) and devstack_services to emit "disable_all_services" so to
cancel any devstack default.
The variables are defined as global ones as well as host-vars for
the controller and group-vars for peer nodes, so that any
descendent job may extend them, thanks for Zuul dict merging.
Change-Id: I2cdb723f6ee209683044fecec59ff7b510a2752b
If user try to create a server in local script before nova cells is
configured, it will run into the following error:
Host 'x' is not mapped to any cell.
Change-Id: I4fe76865fd6e16d5beb5ed9e5d6a9f3542e990a5
The function was introduced in [0] using a hardcoded timeout of 60
seconds which turns out to be too small on slow machines. Create a new
global variable NOVA_READY_TIMEOUT instead so that users can
override the timeout if necessary.
[0] I32eb59b9d6c225a3e93992be3a3b9f4b251d7189
Co-Authored-By: Mohammed Naser <mnaser@vexxhost.com>
Change-Id: I0cd7f193589a1a0776ae76dc30cecefe7ba9e5db
The xen tools have been moved to the project of *os-xenapi* since
os-xenapi 0.3.0. We also did some refact work on these tools in
os-xenapi. This commit is to remove these tools from devstack. So
that os-xenapi will be the single place for xen tools.
Change-Id: I4fdbe6bce12dfedd0d1e975ab8dd624ee3740c11
We've called the jobs that don't run on our main Ubuntu targets
"platform" jobs; start at moving these jobs to native jobs.
Depends-On: https://review.openstack.org/541010
Change-Id: Ib64d91206a9ac677f4d77873bc54c6a84702d6c3
All the evidence from [1] suggests that on opensuse swift-init is not
detaching the daemon process correctly. It's possible there's a pipe
still in play that somehow holds our ansible-streamer open.
This is a minimal fix to avoid swift-init. Although it's possible in
non-default paths to still use swift-init (and hence possibly hit
another variant of this issue), after discussions with swift
developers it was decided the intersection of tests running under our
current ansible, on suse, that would enable these services is
sufficiently small that this is the best course for now.
[1] https://storyboard.openstack.org/#!/story/2001528
Change-Id: I1b68c08c07cf6653ea58506f738cbe0054b38f3a
As described in the documentation, this flag is intended for the case
where the console output is being captured by a tool that appends its
own timestamps.
In the gate this is the job-output.txt. We want the console output as
people like to watch that scrolling by as part of the live console
log. Although this gets saved to job-output.txt, we still want to
keep logging to the individual log files even though it's technically
a duplicate -- in the multinode case the job-output.txt gets
interleaved by all the running nodes; it's much easier to just look at
the individual log files. Also, people are used to it where it is :)
Change-Id: I3486636f1c76139581f6cd9668426f507b7c621d
I started running this with dib where we have pure python3
environments and it failed.
You can't have unbuffered text i/o in python3 for ... reasons? [1]
Changing the file to binary mode works around this. Python3 opens
sys.stdin in text mode, so we need to manually convert the unicode
strings to bytes before we write them to the binary file.
[1] http://bugs.python.org/issue17404
Change-Id: Iebb26f0d3c2347d262cbc10dfd0912840cd05878
Ansible complains:
The task includes an option with an undefined variable. The error
was: 'dict object' has no attribute 'RedHat'
which is just a mismatch on the "Redhat" string
Change-Id: I447038256561740c224c68388fa5b6a068cc8fed
By default stat is pretty verbose, we can set no_log: true to avoid
adding this info to our logs.
Change-Id: Ia18ebfe179443382cc670ffc4363ab037c43bb85
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Round 2 to add no_log to more things. Specific looping over stats, it
is pretty noise in logs and doesn't seem to add any value.
Change-Id: I580171e0061fa331f3ed510713f1ac7a1a6cb5ea
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
With CentOS 7, ansible is expecting to have RedHat as an attribute for
the dict so Discover configurations task fails with an undefined
variable error.
Closes-Bug: #1750573
Change-Id: I5bf9c4057ca9f75d730add9e429d0ef050c6d900
Since blueprint placement-claims in Pike, the Nova FilterScheduler
uses the placement service to make resource allocation 'claims'
before sending the build request to the chosen compute host to
perform the legacy style resource claim. This allows us to safely
scale out the number of scheduler workers when using the FilterScheduler.
The [scheduler]workers option defaults to ncpu if using the
FilterScheduler (which is the default scheduler driver) so to avoid
out of memory issues, we need to set $API_WORKERS scheduler workers
if using the FilterScheduler in devstack.
Depends-On: Ifdcd363d7bc22e73d76d69777483e5aaff4036e3
Change-Id: Ieae234eb5388560b3f66bf60c156a91a8e831bc4
Somehow this feature was lost in the transition
from q-svc to neutron-api. This patch does not
modify the default behavior but allows specifying
the flag to false to prevent devstack from creating
the public and private networks.
Change-Id: I952672496d007552c0c4d83db0d0df9be50326fc
Signed-off-by: Josh <jhershbe@redhat.com>
Neutron functional tests want to use ubuntu cloud archive but it's
not possible to source the fixup_stuff.sh from a neutron CI setup
script. Break it up so that only the UCA portion can be executed
from neutron.
Change-Id: Ie18833bfa30f1789e63cbe9c86f5ece3453f43fb
The block-storage endpoint was added to the catalog, but in the suburl
case it was not added with the /volume suburl. This leads to find it and
attempting to use it but not being able to because it's mis-formed.
Needed-By: https://review.openstack.org/545117
Change-Id: I84721c8ae637417e4b01be9e546ff77c250fc149
One of the steps when we create a new stable branch is to
branch devstack, then update the default branch for most
repos to use the new stable branch for each repo.
This requires making multiple updates throughout stackrc,
and to further complicate things, there are some repo
branch variables for branchless repos that should not be
updated along with the others.
This can be error prone if not fully aware of these
exceptions. To simplify this process a little, this
patch adds two common variables - one that can be set to
the new stable branch name for all of the repos that
should be branched, and one that can be used for all of
the branchless repos to make it explicit that those
values should be left alone. The cycle-trailing repos
have until two weeks after final release to branch, so
also adding another variable for those to make it easy
to update them at a later time, separately from the
other repos.
Change-Id: I82aa19e739eeda3721bac1cb5153ad0bf2d1125a
The nova-next job is being moved from openstack-zuul-jobs
to the nova repo and so we'll remove it's usage from
project-config, therefore we need to define it's usage for
devstack here.
The related project-config change is:
I36d96f89b3e5323746fcbcef5cc7e4d0384a184d
Depends-On: I24a5f73c29094a23e2fdef8ee8b43601300af593
Change-Id: I28971dc7e5cb0b5cf9698e5251a7bb099e63f3db
In Change-Id Ia3843818014f7c6c7526ef3aa9676bbddb8a85ca the 'host'
setting used for each of the fake compute hosts was accidentally named
'nhost' ('\nhost' was edited poorly), so the setting doesn't actually do
anything: you create multiple nova-compute processes they think they are
all on the same host and only one hypervisor and resource provider is
created.
With the correction in place, the wait_for_compute function needs to be
updated to be aware of the fact that the hostnames on the compute
services will have a numeric prefix when the fake virt driver is used.
Change-Id: I5e8430d170c0b1c4f195ebe510aff8be59e4a3bc
In the run devstack role we specify a path to devstack_early_log then
hand it to stack.sh as a parameter which stack.sh does nothing with.
While looking at a fix for this it was pointed out that these early logs
make it into the job's output log now so we don't need a special file
for them. Rather than handle this as a special case just let the
job-output.txt log file pick up the logs for us which allows us to
remove this unneeded feature.
Change-Id: I9bedbe91c60257d94173b1c70676dd6c2b49dc91
This change mimics how fixed_key would actually be deployed in a real
world environment, with a single key shared across Nova and Cinder
across all hosts.
Change-Id: I50a48e2da57a1cc1ecd250150ea6e9c3745baaca
Move extensions_to_txt to the job defintion so that it may be
extended by descendant jobs.
Depends-on: https://review.openstack.org/540485/
Change-Id: I6e9009faa1451698ed781ce1ffdd9f22c97daa93
The role is logging a lot of useless data with all the stat info
from various config and log files. Remove verbosity using no_log.
Change-Id: I72c721573ffc4a14adc3e2b29285c1071b7ec4f7
Zuul no longer requires the project-name for in-repo configuration.
Omitting it makes forking or renaming projects easier.
Change-Id: I32868cec22149ec1c18fe2737a65e88d32bff531
In some environments it might be useful to resolve the hostname to
an IP address that is reachable by other hosts in the local network.
If some of the configuration scripts have taken care of putting
a proper line in /etc/hosts, do not try to override it by appending
a 127.0.0.1 one.
Change-Id: I7ae20a66c473b0c683803cc44654cd95fcce3639
Closes-Bug: 1746751
The zuul_copy_output variable is designed to be able to be used as
a zuul job variable so that zuul dictionary merging will work. However,
it's currently being set in the playbook rather than as a job variable,
so it's not possible to supplement it in a child job.
Move it to be a job variable. Also remove the wrapping {} as they should
not be needed to make zuul_copy_output a dictionary.
Change-Id: I78c7fed47c2ab868384c74dbff7904d33d510dd9
As libpcre3-dev&pcre-devel are now added as general prerequisite,
remove it from horizon prerequisite.
Change-Id: I872aec210028373c39baee0ab846469fd9920de9
This patch add libpcre3-dev and pcre-devel package for python-pcre
installation.
Closes-Bug: #1745606
Change-Id: I59fc688519341c90dc33b79d536f0625a6c4dd17
In function 'get_instance_ip', 'nova' client command is used to get
instance information in order to retrive IP address of the instance.
There is no need to use the nova command, since 'openstack' client
already supports such basic operation.
Moreover, 'openstack' client has an option to get value of specified
column. That brings more accurate way of retriving IP address.
This patch replaces nova command in 'get_instance_ip' by 'openstack'
command. Note, this nova command is the only one in devstack tree.
Change-Id: Iee0b81a994a4da5b3f4572c2e8eb30514cd43f89
Signed-off-by: Ryota MIBU <r-mibu@cq.jp.nec.com>
Castellan switches the `api_class` config option to `backend`. The change
is still backwards compatible with the old `api_class` setting, but cinder
already updated to use the new option (see I5e46c738531d5d56777e91a00f4cee9531356f2e)
and it is better to use the new setting.
Change-Id: Ib609c82e7076d19676baaf4f08abd79ea11db0e3
There is no way to upgrade ironic before nova because of
grenade design. In multinode job we do not restart nova
as we test partial upgrade of ironic there.
On slow nodes upgrading ironic takes time and nova looses
ironic connectivity
This patch increases api_retry_interval and api_max_retries
to make sure we have a time to upgrade ironic before nova
compute stuck.
Change-Id: I3b1429d6561431a82edda04a0e574cac38771837
There is a hack here to set up ubuntu cloud archive, pinning it to
mirror.dfw.rax.openstack.org. The mirror-info role seems to be doing
this correctly now though, so let's remove the hack and let things work
normally.
Change-Id: I283cb3452245b64e9492806f06404b484f21c358
This adds the ENABLE_VOLUME_MULTIATTACH flag and if True
configures Tempest to run volume multiattach tests.
Note that due to https://bugzilla.redhat.com/show_bug.cgi?id=1378242
we can't run multiattach tests with the Pike UCA packages since
those include qemu 2.10 and libvirt 3.6, and the valid versions for
multiattach support with libvirt is qemu<2.10 or libvirt>=3.10.
Depends-On: I80c20914c03d7371e798ca3567c37307a0d54aaa
Depends-On: I158c6f20e3e6a24bd2e5299abbeb3fc5208e5885
Part of nova blueprint multi-attach-volume
Change-Id: I46b7eabf6a28f230666f6933a087f73cb4408348
Allow users to auto-create a neutron non-flat providernet public network
and use it for external router interfaces. By default, keep the existing
flat network type behavior.
Change-Id: I64f71b0c9fcac97b9b84b7d30ee61659b2a690f1
We have already established the correct version of Python to use during
installation, either automatically or through user-provided information
(USE_PYTHON3, PYTHON3_VERSION). Don't do it again.
Change-Id: I7bdf2be9a885994bf2c437dd104048a1ff2f6666
Closes-Bug: #1744096
This follows a change made to devstack-gate in commit 841ebc3 to allow
tempest to succeed even if it happens to run several volume tests in
parallel. Right now it's possible for a tempest-full test (run without
devstack-gate) to fail with an "Insufficient free virtual space" error in
the cinder-scheduler log.
Suggested by: Clark Boylan <clark.boylan@gmail.com>
Closes-Bug: 1743597
Change-Id: I16ccb9976d1bc7c9f56a6a4d73e35042a5867ef9
CentOS tests have reverted to using upstream for EPEL rather than
local mirrors, introducing some unnecessary instability. The root of
the problem is that /etc/nodepool/provider disappeared with zuulv3, so
we now always re-install the EPEL repo and overwrite the local EPEL
.repos files that were made during test setup and point to local
mirrors.
The other change is that we stopped installing the RDO repositories on
the testing nodes too. That we were incorrectly taking this path and
reinstalling EPEL has hidden the removal of these packages from the
base image in the test, since it ends up installing them too.
Split the install into two parts -- epel and RDO. Check for
/etc/ci/mirror_info.sh (the sourcable mirror script provided by base
test setup) and if so, just enable EPEL so we get the CI-mirror
version correctly. Install the RDO repositories (if not already
installed) unconditionally.
Change-Id: Iccb045a6695deb10da4d68a5694e1fa45ccbb810
'deactivate_image' feature flag was added long back during kilo
cycle. Tempest is going to remove this feature flag.
Depends-On: I843d4c64f24407d9d217005d5ea59d50d7ad62e7
Change-Id: I1ae8efc0e62acc5e05c1c00dc8970b74d8b16da0
Removing the (f23,)f24 support they are EOL.
The only non-trivial change is the apache-httpd default worker change,
however might not be bad idea to use `event` instead of `worker`
in the future, but for now keep it AS-IS and continue to use `worker`.
Change-Id: I96d414a30b58bc4b43da45066fdf310a6a830079
Closes-Bug: #1740194
Added NEUTRON_DISTRIBUTED_ROUTING to more easily control DVR
configuration. If set to True, DVR will be enabled and the
default agent mode will be set to 'dvr_snat' since that works
with all types of routers by default. Advanced users can
override that by setting NEUTRON_DVR_MODE, for example in
multi-node configurations where different agent modes are
desired.
This should bring lib/neutron inline with lib/neutron-legacy
in supporting all the different DVR modes.
Change-Id: I9f25921eefc5b935aad3bb1edc5e41ee0ce43a84
The subnet-range parameter is only sent now if a valid value exists so
the command will not fail
Change-Id: I5296f5b59bc6d3d3db90a685a8678db9a156eece
Closes-Bug: #1718111
It is possibe for the user to override te etcd version, thus download a
different etcd file, but the checksum is constant, so the checksum
verification will fail in that case.
Added the ability to specify a different checksum, so the user would be
able to specify the new version checksum
Change-Id: I85af3af841ae957964f18d4e37a86ab0703882bc
Closes-Bug: #1736718
Worlddumping on a system without net-tools package gets an error showing
"arp: not found". As iproute2 can also show arp tables, we use it
instead.
Change-Id: I0cd83e6d14959dc5a1147c487b11f27fb92aa20a
These files are quite large and disk space is limited so make sure we
compress the log files before copying them to storage. Additionally
os-loganalyze will only operate on gzipped log files so this should fix
os-loganalyze with tempest-full job's logs.
This is mostly a check to confirm everything works as expected but we
probably want to move the gzip step into the log publication roles so
that all log files end up compressed.
Change-Id: Ie87962428e0ca755c211cc5e664a14a9f2a79ac4
In Queens cycle, the code of django_openstack_auth was merged into
the horizon repository. The master branch of django_openstack_auth
will be retired. (horizon blueprint merge-openstack-auth)
This commit drops django_openstack_auth related code from DevStack.
_prepare_message_catalog_compilation in lib/horizon was used only
in install_django_openstack_auth, so it is dropped too.
Change-Id: If9467c520a1e07d1968b29e485df0097330356bc
Do not run devstack and its siblings if only rst files, releasenotes, or
files in doc directory change.
This is the minimal set of irrelevant files shared by most projects
already.
Needed-By: Ie8504ba3d5d46f6338a228ed2d248ba6363e37ae
Change-Id: Id0095763eb91592c2fd1a913526883987df704bd
Nova now calculates alternate hosts in the scheduler and
sends those to the cell for reschedules in case a build on
a given compute node fails.
The cell conductor needs to claim resources against the
alternate hosts in Placement during a reschedule, therefore
it needs to be configured to talk to the placement service.
Part of blueprint return-alternate-hosts
Change-Id: Ie599968d9e7537e551fe6d9deb63a91b256b1e11
This patch creates a new config file glance-image-import.conf
at /etc/glance path. Also, each config option is initialized
with default values.
Need these changes to implement specs [1]:
[1]: https://blueprints.launchpad.net/glance/+spec/inject-automatic-metadata
Related-Change-Id: If14c7dc4f38360006f9cb350fbba54fa2f33be61
Change-Id: I665507db1838a50e344d3be909d7490f1f52040c
fetch-testr-output and fetch-stestr-output are being merged.
Change-Id: I00d448c4e6b98a1f504b048c74eff4e110c0b511
Depends-On: I833320cf9a932d8e119645eb798ce0c93d854321
The file won't be listed as long as it is called .stackenv.txt.gz, with
this it will be called _stackenv.txt.gz instead.
Change-Id: Ib3b44c287ffb2ec0e48fefef1662a1c02d162657
The whole devstack log is written into the console output
(job-output.txt) and into devstacklog.txt.
Remove it from job-output and add a devstack-early log file (same as in
legacy job) that includes all the output of stack.sh.
Make sure the log file is pulled into the stage folder so that it will
end up on logs.o.o.
Change-Id: Ia7c1d8fe5cc03d15f455c6e62ebf4a5f6d62ab1f
Historically we have collected devstack logs under /opt/stack.
Stop doing that and collect them in the stage_dir instead, so that
once the base job logs pull service comes around we are ready for it.
This add the benefit of writing things into a folder which is
already owned by the ansible user (ansible_user_dir), so we don't
run into issue writing there.
A few logs (devstack log, log summary and dstat) use to show up on
logs.o.o. just because they happened to already be in /opt/stack/logs.
With this change they would be lost, so adding them to post.yaml.
Depends-on: I5ad4dfccbc1389da3afc53f3c866d3475e006db6
Change-Id: Ib4be2f5056c0dc2b776de4a0d18b47b12624be92
The sphinx jobs need to find doc requirements in either
test-requiremnts.txt or doc/requirements.txt. Putting them directly in
to tox.ini, not so much.
Change-Id: I98a43b511a6949fa4f00c26eec224d24d6fa6588
Tempest is going to test volume v3 APIs as default
in gate and running a separate job to run tests on v2 APIs.
To give this ability, this commit provide a var to tell
which API version need to be tested and accordingly it
configure the catalog_type and microversion setting on tempest.
Change-Id: I531f3b32e81ac5d282461597ca286c09429cb143
Needed-By: I0c9193501eb9eaa25eb5f0786bb72eb7855099fb
The loopback device is created for ceph osd. If the directory
${storage_data_dir} has been mounted when create disk, we should
umount ${storage_data_dir} instead of ${storage_data_dir}/drives/sdb1.
Change-Id: Ie9fe81c820c485dab9f049cf5a81c02424925728
Closes-Bug: #1689089
The current Glance config files are a combination of copied and
generated files. This patch makes all the files generated and
removes now unnecssary ini(un)comment statements. It additionally
removes some ini(un)comment statements that weren't having any
effect on the previously generated files.
Change-Id: I6e4b7694e8bebb7fe6661ead034ee257c768e342
I keep copy-pasting these to projects from the shade repo. Let's make
some base jobs people can more easily use.
devstack-tox-functional runs devstack and a tox functional environment.
devstack-tox-functional-consumer is the same, but runs devstack in pre.
It's intended for projects for whom patches to the project won't
actually impact the devstack deployment (shade, nodepool, gophercloud
are all examples of such things)
Change-Id: I84de60181cb88574e341ff83cd4857cce241f2dd
Without installing the targetcli package
tools and configs can be missing.
The code was correct baside a typo,
it is `ISCSI` not `ICSI`
Change-Id: I32e5d84d87560458f0eaaf820dcd00c86e6dec8b
The devstack base job in in use in many projects, but it is not being
gated here in devstack. Let's add it to the list so that we don't
accidentally break it.
Change-Id: Iea13235a8438d4b540f9f27b94aed13e719481dc
Use the test-matrix role from devstack-gate to define a base set of
services to be enabled for the controller and compute nodes.
Extend the local conf module to handle the base set of services.
Since the test-matrix defines services for primary and subnode nodes, we
need a multinode job to test that this works. Add a new host group
called subnode that includes the non-controller hosts. Add a new job
that runs devstack on a two nodes environment.
Using service from the test matrix enables swift in the gate, so we need
to set SWIFT_HASH for devstack to work.
Depends-on: Ie36ba0cd7cfcd450b75000a76a64d856f2a83eba
Depends-on: Id9ad3be4be25e699f77d6b5a252f046ce8234f45
Change-Id: I379abf482c89122533324e64fefbff3d5a618a89
It had been in stop_placement, but we don't want it there: the old
side of grenade needs to call that but should not remove the uwsgi
configuration when doing so. It is configuration, after all.
Change-Id: Iee763adf7895145d97b184924896db3f1f48a015
Partial-Bug: #1736385
Several legacy jobs use the OVERRIDE_ENABLED_SERVICES variable
from d-g so set the list of services that should be enabled and
ignore the default set calculated via the feature matrix.
Add support for a similar functionality in the zuulv3 jobs
using the 'disable_all_services' localconf function.
Change-Id: I690554ec62cef3be600054071efbb3f92a99249e
python-openstacksdk does not match its pip name which is openstacksdk.
So setting python-openstacksdk in LIBS_FROM_GIT leads to devstack
thinking there is a problem.
Put in a workaround for now. It would be better to either:
a) rename python-openstacksdk repo to openstacksdk
b) rename the pip name for openstacksdk back to python-openstacksdk
c) add general support in the various GIT hashes for a pip name
Change-Id: I57cf95763d54ad2060a4ce2af91c3ba18ca04db0
Cinder has now implemented "policy in code" and policy.json is
only needed for overriding default policies. The default policy.json
file has been removed in Cinder so we need to stop trying to copy
it during Cinder setup.
Change-Id: I364e401227fe43e2bacf8a799e10286ee445f835
This will enable us to run one etcd for devstack and another
for kubernetes in the same box if necessary
Change-Id: Ib71ded24727b80afd4d98eb68bade0f8c0f72311
The local requirements repo can be checked out to a stable branch,
in which case, the requirements might conflict with tempest's master
requirements.
Master branch's upper-constraints should be used when building tempest's
venv.
Closes-Bug: #1706009
Change-Id: Ifd64638cae2886671421149dbbff3a57f9c64257
The 'pip list' command prints the "safe name" which converts _'s to
-'s amongst other things; e.g. glance_store becomes
"glance-store 0.21.1.dev22 /opt/stack/glance_store"
Because people may use these more familiar "file system" names in
LIBS_FROM_GIT automatically convert names when checking if libraries
are installed.
Change-Id: I30524f80a341f38dfa794a8f629d859e85a4a448
In commit f0cd9a8b08 we changed to
use column format, but it checks for zero length string and
check_libs_from_git fails.
Change-Id: I97b52b80efb33749647229a55147a08afa112dd2
Add a no-op function, "plugin_requires" to allow plugins to indicate
their dependencies on each other. This will be used by the Devstack
Ansible module when writing local.conf files.
Also add define_plugin to allow plugins to indicate their canonical
names.
Change-Id: Ibd8c7222ed7dfb08d7ea821d871fc6f3b88de24b
Some old configuration(such as, LOG_COLOR config) will remain
if we don't cleanup.
So, we should cleanup the configuration before we config it.
Change-Id: I7aff609dadf3acba13a36894614b35005f51280d
The kvmibm removal I009ae4779588615633bff81d0c47a1b879ec9279
incorrectly removed this (the check was install if *not* kvmibm).
Since we don't support kvmibm any more, it should be safe to install
everywhere as done here.
For the full history, it started with us installing qemu-kvm-ev with
Ide91b261f35fb19d8bd7155ca016fa3b76a45ea1, then we fixed it to be more
generic and just install qemu-kvm with
I46da627c0da8925064862fdc283db81591979285, then Fedora 26 support in
I5c79ad1ef0b11dba30c931a59786f9eb7e7f8587 made this install everywhere
*but* kvmibm.
Change-Id: If3e9661451ad1055e7c8d670605a53095f0aeda4
With the release of 17.10(artful), support for 16.10(yakkety) has ended.
Update our list of supported distros accordingly.
Change-Id: Id85e00f109cfd43141dec0c0d2bfedb66f14e664
This commit switches TimeoutStopSec in DevStack's systemd unit files
from "infinity" to "300". There are two motivations for that change:
* 5 minutes should be more than enough to stop a service.
* systemd included in CentOS 7 and RHEL 7 doesn't support "infinity" as
a value, "0" should be provided instead. When "infinity" is set,
systemd will kill the service instantly, leaving service children
processes orphaned. Instead of differentiating here, we can just set a
sane, finite number.
Closes-Bug: 1731275
Change-Id: I0a079ea9879fa4fbba23104c2f5ab6e0721a2a2a
Option nova_metadata_ip was deprecated in favor
of nova_metadata_host. lib/neutron was updated
recently but lib/neutron-legacy was missed.
Change-Id: Iadd42458dda705ad0c24aa4ab2afd5b27dd8f0e1
Currently doing a cycle of
./stack.sh; ./unstack.sh; ./stack.sh
fails because the leftover tls-proxy sites will cause apache startup to
fail on the second stack.sh run. So we need to disable these sites on
running stop_tls_proxy.
Change-Id: I03e6879be332289d19ca6a656f5f9f139dffff6f
Closes-Bug: 1718189
The mysql-community-server is a compat provide, openSUSE uses
mariadb for quite some time. Make it futureproof in case
the compat provide goes away in the future. Cleanup
mysql service name to MYSQL_SERVICE_NAME and consistently
use it.
Change-Id: I2df7b8d8b798dfa7ceade90e0c127e0609524a8b
Zuul now supports including the file extension on the playbook path
and omitting the extension is now deprecrated. Update references
to include the extension.
Change-Id: I4bff5f12742364f7cc92e17869a047fd2185dda4
A proposed change[1] to Zuul removes the implied run attribute.
Add an explicit run attribute here to prepare for that.
[1] Ia8f23bce9898cd4f387554e6787b091b63e75519
Change-Id: I1fbc36c3d1b8c4ed70fceef1c587255dad50da04
* Since ENABLE_KSM param will be used in local.conf file
and it's value is received in a variable and while compairing,
the variable needs to be compared. So we need to change the
same.
Change-Id: Id4ed17c0642acd2313e456503cfc375ca6f61409
Closes-Bug: #1724690
The old code was strip()ing the version string instead of split()ing the
version string so we always got the first character of the version
string. This worked fine as long as the pip version was single digit but
as soon as it rolls over to '10.stuff' we will compare:
pip version 1 (instead of 10) > 6
Which fails bceause 1 is less than six. Instaed we really do want to
compare 10 > 6 so use split on '.' instead.
Change-Id: Ic7d0c04d7fa77774ab2d70fb9d11f182becec553
Nova is gaining the ability to run TLS over the connection between the
novnc proxy service and the QEMU/KVM compute node VNC server.
This adds a new config param - 'NOVA_CONSOLE_PROXY_COMPUTE_TLS=True' -
which instructs devstack to configure libvirt/QEMU to enable TLS for the
VNC server, and to configure the novncproxy to use TLS when connecting.
NB this use of TLS is distinct from use of TLS for the public facing API
controlled by USE_SSL, they can be enabled independently.
This is done in a generic manner so that it is easy to extend to cover
use of TLS with the SPICE and serial console proxy services too.
Change-Id: Ib29d3f5f18533115b9c51e27b373e92fc0a28d1a
Depends-on: I9cc9a380500715e60bd05aa5c29ee46bc6f8d6c2
Implements bp: websocket-proxy-to-host-security
The pip list command supports a --format=columns option which outputs
things in space delimited columns. Switch to using that.
Change-Id: I5140a7d83bf567b1c3c67516112eb4c57074fa53
Reason is to be identical to the upstream KVM CI. Some Tempest
tests rely on vdX virtio-blk device naming. Others simply create
their own with a brand new image. Also, the scsi support on the
CirrOS image is limited, tests booting from volume fail.
Change-Id: I389147a58042aa6098a695e6dd32f3e697fbbbab
This makes the condition that chooses which daemon name libvirt to call
the same as for choosing the livirt package names.
Without this fix the condition checking for a directory is incorrect
when livirt is not yet installed, but is used before installing the
packages.
Change-Id: Ib5eb12769128527a6f4b3b5f7674bd2dad0ed160
VPNaaS agent is going to be an L3 agent extention.
Related-Bug: #1692128
Depends-On: I0b86c432e4b2210e5f2a73a7e3ba16d10467f0f2
Change-Id: Id827274b7c74cdf71db6d1f2ab3eadb5fef099f5
The file matcher was from the early versions of this when we were
running both v2 and v3. We should always run the new devstack
job on all changes to devstack now that v3 is in production and we
plan on building jobs off of this one.
Change-Id: I7dd336b0059043f6653bdfdcba0ee5cded3e67b1
This should be managed in the devstack repo, since it's a base job to
run devstack.
Change-Id: Iffe54fbccbccd68db08f79a1b51dd7f76dbff408
Depends-On: Ie2119f24360d56690ffd772b95a9ea6b98dd4a39
As the following commit has renamed the two vnc options; let's
use the new options in devstack:
https://review.openstack.org/#/c/498387/
Change-Id: Id125666814ea9bb8a22b579aee0f6bc1c65ade80
block-storage is the official service type for cinder, according to the
service-types-authority. Add it as a service in devstack, with cinder's
unversioned endpoint, to enable proper discovery.
Change-Id: I75cf7212678f7f270c3c32f0bce227dbbf6b466d
The IBM hypervisor distro "KVM for IBM z Systems" gets discontiued,
like announced in March 2017 [1]. The key dates are:
* 03/2017: announcement
* 08/2017: the last day to order (EOM)
* 03/2018: the End of Service (EOL)
As the CI which tests OpenStack with KVM on IBM Z doesn't rely on this
distro anymore and EOM has reached, we remove the Devstack support for
this distro.
This basically reverts commit a5ea08b of Dec 2015.
NOTE: This doesn't affect other distros which have KVM on Z support.
References:
[1] FAQ for KVM for IBM z Systems Delivery Strategy Change
https://www-01.ibm.com/common/ssi/cgi-bin/ssialias?htmlfid=ZSQ03110USEN&
Change-Id: I009ae4779588615633bff81d0c47a1b879ec9279
Strip the [<extras>] string from a <package_dir>[<extras>] argument
when looking for the package directory. Explain what the heck is
going on.
Change-Id: I79beb5c3e9e7c35c91cdd0d5a1d91532bebc4b6d
Closes-Bug: #1721638
We install the glance api on all nodes in multinode testing. This has
been failing because we don't configure the glance auth cache dirs if
we only install the glance api service. This was done as part of
init_glance which is only run when installing g-reg.
Fix this by moving the auth cache dir creation step into
configure_glance which is run for the glance api.
Change-Id: Ie669827507df0f524e6e53fe4ab3dff848dd4bd7
This reverts commit ef5ebed6c9.
The problem here is a backwards-incompatible change to
configure_auth_token_middleware. Plugins are still passing a
"signing_dir" which is interpreted now as the "section" argument
... this leads to an interesting red-herring issue; because "v" is a
gnu sed command for checking the version, a signing_dir of "/var/..."
(as done in most plugins) gives the weird error:
sed: -e expression #1, char 32: expected newer version of sed
I think we'll either need a new function, or dummy arguments to get
this back in.
Change-Id: I2098d4eb2747282622cf486fa7dbf216f932f58b
As described in the change, "pip freeze" has issues with the way
zuulv3 clones repos without a remote. This is an attempt to use "pip
list" to check for local install
Change-Id: I33d25f86b6afcadb4b190a0f6c53311111c64521
"pip freeze" reports an error when checking because the repos don't
have a remote under zuulv3
---
Error when trying to get requirement for VCS system Command "git
config --get-regexp remote\..*\.url" failed with error code 1 in
/opt/stack/new/keystone, falling back to uneditable format
Could not determine repository location of /opt/stack/new/keystone
Complete output from command git config --get-regexp remote\..*\.url:
---
This means this check fails. I think we can fix this by looking at
"pip list" which I will propose in a follow-on, but this fixes the
immediate breaking issue.
Depends-On: Ib12ddf768ee20fd7614622179f6842f5d57864ff
Change-Id: I21ff749ab3e7911fa074e6d53056768f42f8aa57
Change f119121d21 removed
the default image to download which meant if you were using
the fake virt driver, no image would get downloaded and
tempest setup would fail.
This adds it back in but doesn't use a wildcard.
The default image is the same as before, but uses the
variables that are also used for the default libvirt image
case.
Change-Id: I80eddd0d3a99572ed494b5cd36fed8ceb4d05d77
Closes-Bug: #1720003
Memory_tracker imports psutil, but does not run inside a pip/virtualenv
so the system provided psutil library needs to be provided. This
is matching what is done for other non-SUSE distributions
Change-Id: I96f944730dc8644333d906d71339351b29b03e08
PKI tokens have been actively deprecated from keystone and there are
deprecations being emitted from keystonemiddleware. Because of this we
no longer need an auth cache directory in the services where the PKI
certifcates used to be stored.
Remove the creation and use of all these AUTH_CACHE directories.
Change-Id: I5680376e70e74882e9fdb87ee1b95d5f40570ad7
I had to dig around for awhile to figure this out, so
this adds an example on how to grep journalctl nova logs
for a server instance UUID.
Change-Id: I6a5c47fbcba3af1822e2f9efc2ac20ebe0387f3f
Castellan switched the `api_class` config option to `backend` in commit
8980bf7da55dd084ad84c84534fe937f0d43b9c0. The old setting will still be
recognized for now, but we should switch to using the new, correct config
option.
Change-Id: I5e46c738531d5d56777e91a00f4cee9531356f2e
We want that one to be installed via pip, if we still use it
(by default PyMySQL is used, which is already installed via
pip as well).
Change-Id: I76454aa7f84379aa387b144686bcfaa327b141ed
There have been a couple of new stable releases in the meantime, update
to using v3.1.10 which is the currently latest stable version.
Change-Id: Ifa1421c9f12af9753052f992929deb7ebd45e804
lsb-release is a dependency of "lsb", so it used to work
before just fine as well, but it was installing about 300MB
of "stuff" that we don't actually need..
Change-Id: I25c7c750cbaeb40bf4f2e8695608c4b1003289ea
In a devstack environment you likely need to use sudo
to run the journalctl command, so this adds that to
the examples.
Change-Id: Ibe6b71285a3014e80e06a50130f18bfbdb4ff3ab
The v2.0 identity API is being removed in the Queens release, but in
order to do so we need to stop some v2.0 tempests tests from being
run. This commit switches the default to disable the keystone v2 api.
In a future commit after the removal of the api from keystone the bits
to deploy the v2 api will be removed.
Change-Id: I5afcba6321f496b8170be27789bee7c9ad8eacce
(1) when checksum fails, better delete the broken files and try the second time;
(2) amazon s3 is not good in mainland China, better try one more time with wget
Change-Id: I24ee73f216b78bd80564863cd335e5d5a9b56360
Otherwise neutron will fail to bind external ports because of missing
entries for external physical network in the mapping.
Configure it only when l3 agent is also installed on the node (otherwise
the l2 agent is not exposed to external network and hence doesn't have
the bridge).
Change-Id: I561b74538acb0dc39f1af3e832108ce6a99441b0
The client are told to connect to SERVICE_HOST instead of HOST_IP, so
we need to start etcd3 with matching listening parameters.
Change-Id: I96389090180d21d25d72df8f9e8905b850bcaee9
Partial-Bug: 1656329
This reverts commit a7e9a5d447.
The jobs that run live migration tests are failing at about
a rate of 50% since this merged. There are no recent changes
to nova in the last 24 hours that are related to live
migration, and this is failing on the master branch only,
so I suspect the failures are due to new qemu packages
getting pulled in from this change.
Change-Id: Ic8481539c6a0cc7af08a736a625b672979435908
Closes-Bug: #1718295
By default memcached is bound to 127.0.0.1 and we have no code in place
to change that. So instead of using the $SERVICE_HOST variable, we
hardcode it to localhost, just as we do for the cache settings, see [1].
This also avoids a bug that occurs when $SERVICE_HOST contains an IPv6
address, as in that case it would have to be prefixed by "inet6:" [2].
[1] I95d798d122e2a95e27eb1d2c4e786c3cd844440b
[2] https://bugs.launchpad.net/swift/+bug/1610064
Change-Id: I46bed8a048f4b0d669dfc65b28ddeb36963553e0
Partial-Bug: 1656329
In the "DevStack Component Timing" section, display the unaccounted
time.
Also add the units (seconds) to the output to make it clear to viewers.
Change-Id: Iaca82cc54a355f7077e20e548b771e53387f6628
https://review.openstack.org/504171 prevented Python version detection
from failing when python3 is not installed. However, "which python3"
returns a message in stderr when python3 is not there, and this output
can make diskimage-builder get confused when parsing
source-repository-images.
Change-Id: Idb649dc341ede73c39954b0432ef3cacf379ed37
On platforms without python3 installed devstack fails the python
version check. This does it more gracefully.
Change-Id: I4d79a41eb2d66852ab1a1aa8bd383f8e3e89cdc0
lib/neutron service prefixes are neutron-* not q-*. We should install
those packages either way.
The patch moves files/*/neutron into files/*/neutron-common so that we
can correctly match */neutron against service specific dependency files
(f.e. */neutron-agent) and load the common packages if any neutron-*
service is present.
Change-Id: I57b36f2ed3f33737223a35d9ed734bb414f31e0b
Since [1] devstack is failing on s390x with "Distro not supported".
The reason for this is the missing etcd3 support. It worked before
[1] as we were able to disable etcd3 via local.conf. But as etcd3 is
a base service, we might not be able to rely on this workarond in
the future anymore.
As there is no etcd3 binary hosted on github like it is for other
architectures, the user needs to specify an alternative download
URL via local.conf. Otherwise devstack will exit with an appropriate
error message.
ETCD_DOWNLOAD_URL=<custom-download-url>
[1] https://github.com/openstack-dev/devstack/commit/d8bb220606737719bcdf7c5b4f54906f2974c71c
Change-Id: I1c378a0456dcf2e94d79a02de9d3e16753d946d6
Partial-Bug: #1693192
DVR isn't supported by the Linux Bridge agent, but the
mechanism driver is enabled by default, so Neutron attempts
port-bindings for it, generating ERRORS in the neutron-server
log in the check and gate jobs. Just remove it in the DVR case.
Change-Id: Ic50e12e5fecf366a182c141b5c99649e653254cb
Closes-bug: #1716782
Without this patch, the pmap -XX call fails on openSUSE Leap
distributions as those have a fairly ancient procps version that
does not support the -XX parameter. A pure python implementation
is more portable, faster and even shorter than the subprocess
call.
Closes-Bug: #1716066
Change-Id: I2fdb457e65359a1c9d40452c922cfdca0e6e74dc
uwsgi services:
[1] By default uwsgi is set to exit on reload this breaks config reloading
of the service [1][2]. It needs to be set to 'false'.
[2] Requires to add 'systemctl reload' command support by adding ExecReload
in unit file.
Non uwsgi services:
[1] Non uwsgi services only requires to add ExecReload in unit file.
There was a similar patch submitted by Matthew Treinish [3] but it was
already set to workflow +1(not merged as having dependency on other patch)
and it was having some issues as specified in comment and missing reload
functionality for other services.
[1] https://etherpad.openstack.org/p/uwsgi-issues
[2] http://uwsgi-docs.readthedocs.io/en/latest/Options.html#exit-on-reload
[3] https://review.openstack.org/#/c/490904/2
Change-Id: I78f5e9d4574671c74a52af64724946feb41c2d7a
Add instructions for installing and enabling remote-pdb [1] under
systemd.
[1] https://pypi.python.org/pypi/remote-pdb
Thanks to clarkb for pointing me to this. TIL.
Change-Id: I640ac36cfbcc5b199e911c0e3f6b18705c3fbbc4
This makes sure that the openstack client placement plugin gets
installed when either LIBS_FROM_GIT includes osc-placement or
placement is used, which is always now if you've enabled nova,
which is enabled by default.
Co-Authored-By: Matt Riedemann <mriedem.os@gmail.com>
Depends-On: Ica83e29780722dd1555904e46b9ff1d1fdf72516
Change-Id: I7c5a5c10288b356302bb3027837d4ed4f0fbad8c
This was previously set to thin as the default, but at the time
there were failures seen with what appeared to be race conditions
when creating snapshots.
These failures are not seen locally, and we have a lot of installs
using the default auto by this point with no reports from the field
of seeing this failure. This is to be able to more extensively test
this in the gate, and hopefully get this switched over to be able
to thinly provision by default when possible.
Change-Id: I3e99adadd1c37ba8b24b6cb71a8969ffc93f75a1
Related-bug: #1642111
Now that Pike has been released, switch to using the pike version of
UCA instead of ocata, too.
One reason to do so it that it adds python3-ceph packages, allow to have
progress with the python3 compatibility of the Ceph integration.
Change-Id: I7d95e53892b697c72af75ad0ce7ce2dec6d31fde
For devstack master branch we should point to the latest version and not
to some possibly outdated older branch.
Change-Id: I4af3aef90a2c295df3de4a5b49d127e85ab517ac
1] Process using uwsgi:
uwsgi services doesn't support for graceful shutting down [1].
It requires some changes in unit files [2] including adding below
graceful shutdown hook and changing KillSignal:
--hook-master-start "unix_signal:15 gracefully_kill_them_all
All the steps and changes required are specified in etherpad [1].
2] Non uwsgi services needs below changes:
In [service] section:
a. Add KillMode = process
b. Add TimeoutStopSec = infinity
NOTE:
Creating unit file for services other than uwsgi is handled by the
'write_user_unit_file' function [3]. This function is common for all
the services so this patch adds the above mentioned parameters for
services using ServiceLauncher also though they don't require.
Added a new stackrc variable WORKER_TIMEOUT which is required to add
graceful shutdown support to uwsgi services. It will be set as a value
to 'worker-reload-mercy' [4] in uwsgi file of service. The default
value set to this variable is 90.
[1] https://etherpad.openstack.org/p/uwsgi-issues
[2] https://www.freedesktop.org/software/systemd/man/systemd.kill.html
[3] https://github.com/openstack-dev/devstack/blob/2967ca3dfd0d64970dfa5dfa7ac2330ee7aa90ed/functions-common#L1439-L1461
[4] http://uwsgi-docs.readthedocs.io/en/latest/Options.html#worker-reload-mercy
Co-Authored-By: Dinesh Bhor <dinesh.bhor@nttdata.com>
Change-Id: Ia95291325ce4858b47102dd49504250183f339ab
We need this for every devstack run now, so downloading it from github
every time isn't the most awesome thing in the world.
Add an extra variable EXTRA_CACHE_URLS which will be appended to the
output of tools/image_list.sh. This way, these files will be
downloaded during the daily nodepool build, but they will not be in
the IMAGE_LIST and hence be considered as images to upload.
Add a function get_extra_file which echos the path to a file given the
URL. It will first check the cache at $FILES, and if not present
download it.
Update the documentation in image_list.sh to reflect what's happening.
Move the defaults for etcd variables into stackrc, since it is a base
service now.
Change-Id: I86104824a29d973a6288df1f24b7891feb86267c
1) Update doc links according to OpenStack document migration
2) Use https instead of http for docs links
Change-Id: I81b560d1e4c5210dc00a6a6ac06c03bb1e69d595
This is proposed temporary change to avoid
"ValueError: need more than 1 value to unpack" error in stack.
Change-Id: I743febbef3a1f201cea37471356518be31585277
This tears out the alternative path of using screen, so that we only
use systemd enabled paths. This simplifies the number of ways that
devstack can be run, and provides a much more reliable process
launcher than the screen based approach.
Change-Id: I8c27182f60b0f5310b3a8bf5feb02beb7ffbb26a
Leaving the default user enabled is a security issue, as it can be used
without credentials. It also may mask issues like seen in [1].
[1] https://bugs.launchpad.net/bugs/1651576
Change-Id: I75b4e5696c0f8017b869127a10f3c14e2f8bd121
Various services are returning broken links when running behind
tls-proxy. These issues can be fixed by setting the X-Forwarded-Proto
header in the apache config and letting oslo_middleware parse it.
Change-Id: Ibe5dbdc4644ec812f0435f59319666fc336c195a
Partial-Bug: 1713731
The pre-existing configuration for swift on devstack set's the
*-server's devices option (the root of the servers list of devices) to:
devices = /opt/stack/data/swift/1
where "1" is the node_number, and will be 2, 3, ... N if the devstack
machine is built with more than one swift node/device (pretty sure no
one does that on devstack ever).
The device(s) in the rings are named (perhaps confusingly similar to the
swift loopback image) just "sdb1", so all storage servers expect to have
a $STACK_USER writeable file system at:
os.path.join(<devices_root>, "sdb1")
That directory does not exist when you start up a devstack [1].
Currently Swift's object-server's require that directory exist before
they write data into it (even with mount_check = false!).
Unfortunately however, with mount_check=false the account/container
servers are able to create the device directory when it does not exist
[2]. Which can lead to some unfortunate results with permissions on
some deployments using mount_check = false (e.g. testing or
containerized environments). Fixing this issue [3] uncovered the
previously benign [4] mis-configuration in devstack.
Attempting
1. It was lost a long while ago I7c65303791689523f02e5ae44483a6c50b2eed1e
2. Essentially they want to:
mkdir -p /opt/stack/data/swift/1/sdb1/containers/<part#>
... but end up creating the "sdb1" dir too!
3. I3362a6ebff423016bb367b4b6b322bb41ae08764
4. Benign because the object-server share their device with the
account-container devices and they would create the dirs before trying
to write an object. It was incorrect, but worked by happenstance, which
is nearly as good as worked on purpose.
Change-Id: I52c4ecb70b1ae47e613ba243da5a4d94e5adedf2
In [1] the definition of the function was changed, adding the service
name as first parameter. Since this seems to have caused failures in
some plugins, at least update the function template accordingly.
[1] Ifcba410f5969521e8b3d30f02795541c1661f83a
Change-Id: I4d03957f8d3a18625f06379fb21aa7ba55e32797
Dragonflow can operate without L3 agent, and does not advertise L3 agent
scheduler extension when running this way
Change-Id: I23d0e558c8454636fcde0a1903c78965b70bc324
I am a new participant, and when I read the document I found that the command in the OS_AUTH_URL example is v2.0, so I want to update it.
Change-Id: I973adc303a3cb37ce377ca4e31d1d666cd41b358
When run in the default superconductor mode, the screen-n-cond-cell1
logs are not formatting in oslo format or colorized like the other
logs. This is because screen-n-super-cond is running using nova.conf
which is configured for oslo format logging with color.
The oslo format logging is also needed to correctly index the logs
from screen-n-cond-cell1 in logstash.
This change simply configures nova_cell*.conf files for logging
like nova.conf.
Change-Id: I44fc11f09bb7283be0b068f5e02a424f3e5dafe2
Closes-Bug: #1713070
Query the python2/python3 interpreter for it's version to fill in
PYTHON3_VERSION and PYTHON2_VERSION defaults. This means on a
python3.6 platform such as Fedora 26, we don't need to override the
default.
Change-Id: Id826f275b99b9f397b95e817941019fc503daa1d
Add a function which tests if a plugin has been enabled with
enable_plugin. This is helpful if two co-ordinating projects want to run
specific setup in devstack in one only if the other is enabled.
Change-Id: Ibf113755595b19d028374cdc1c86e19b5170be4f
* Check KEYSTONE_DEPLOY flag and cleanup appropriately
* When we stop process, we should not wipe uwsgi config we should
remove files only on cleanup
* We should not call cleanup *BEFORE* configure, we are just wiping
out the uwsgi ini files
* cleanup_placement should be called from clean.sh
Change-Id: I066f5f87ff22d7da2e3814f8c2de75f2af625d2b
Code in grenade and elsewhere rely on the process/service name
when one runs "ps auxw" and they grep for example "grep -e glance-api"
to check if the service is running. with uwsgi, let us make sure
we use process name prefix so it is easier to spot the services
and be compatible with code elsewhere that relies on this.
Change-Id: I4d1cd223ed9904fcb19b26fc9362b676e0b4f9b3
This adds a new feature flag on tempest conf whenever LDAP is enabled.
When this flag is set to True Tempest users and groups identity tests
adapt to fetch users and groups from different domains.
Change-Id: I368ddf34908b906355c422bd1afd6ab9b1a80053
Depends-On: Iedb470c51fa2174ab7651e6b7e22eff1f25f7aac
We want people creating plugins (that add services) to be aware of the
service-types-authority (STA), so this change adds a Prerequisites
section and notes the existince of the STA there, and the need to apply
there to create a service-type.
Change-Id: I1aa48fe231aaa4499f8b4fe336abea668841b9af
stop_nova_conductor dropped the ball when the CELLSV2_SETUP mode is
set to "singleconductor". We should cleanup the older style "n-cond"
in this case.
Change-Id: I9ffd6d09df6f390a842b8a374097f144564d2db4
... to be used with domain_remap. Swift will start functionally testing
domain_remap in I63428132283986bda9e5c082ffe85741449b71ba.
Change-Id: I4c1ab06d040d91fd8c314d0aa2cecbbb00adf8ad
Now that the pike branch exists we don't want to special case grenade +
system on the target side. We should use systemd for both sides of the
pike -> master upgrade. Note this change should not be backported so
that we do not attempt to use systemd on the ocata -> pike upgrade path.
Depends-On: Iedf824a1772115e0dff287a898636f8e58471269
Change-Id: I6198bf1842a44773fce80672c81eee3afc3c6f38
Calling setenv appears to be globally scoped which is breaking the
glance path which relies on chunked uploads. The glance path is
separated by using mod_proxy instead of mod_proxy_uwsgi because
mod_proxy_uwsgi doesn't support chunked encoding.[1] The proxy-sendcl [2]
was set on the mod_proxy_uwsgi path just in case someone tried to send a
chunked request to the api server we would be able to handle it. It
tells apache to locally cache the chunked request and send the
content-length as a normal upload to the upstream server. However, if we
can only set it globally across then small potential benefit is not worth
having all glance uploads cached by apache. This commit just removes
setting the flag. In the future if we can have devstack isolate this
flag it might be worth adding back to the mod_proxy_uwsgi path, but for
right now it's not worth the tradeoff.
[1] https://github.com/unbit/uwsgi/issues/1540
[2] https://httpd.apache.org/docs/2.4/mod/mod_proxy.html#request-bodies
Depends-On: Idf6b4b891ba31cccbeb53d373b40fce5380cea64
Change-Id: Iab2e2848877fa1497008d18c05b0154892941589
Closes-Bug: #1709970
lib/nova does a pip re-install of libvirt-python to rebuild the python
library incase the underlying libvirt version changed during package
installs. In offline mode, the underlying version of libvirt can't
have changed; so we have the situation that we've removed the libvirt
python bindings but can't reinstall them (because we're offline).
This fixes that particular situation, but skipping uninstalls in
offline mode seems generically OK.
Change-Id: I2b75d45d94d82f87d996c7570c125d46f5f99f6a
Closes-Bug: #1708369
Configuration options that toggle support for LDAP read/write
were deprecated and removed as of the Ocata release:
I13eada3d5c3a166223c3e3ce70b7054eaed1003a
This means we no longer need to clutter the domain-specific
configuration with these values since they are no longer used.
Change-Id: I23b5b994862f066c3d48ce524c396faecabf60f8
In Ocata, we replaced the verification logic for CPU, RAM and disk by calling
the Placement API instead of using those legacy scheduler filters, it's time
to remove them from the default list of filters that are run, especially
since Nova now removes them from the conf opt defaults thanks to
Ibe1cee1cb2642f61a8d6bf9c3f6bbee4f2c2f414
Change-Id: I2e81f1bbce7476d63e84e70dcdd59a1163f89f09
Related-Bug: #1709328
i had sync the all repos of devstack for installing in intranet
environment, and found the url of noVNC had changed to
https://github.com/novnc/noVNC. This module upgrade from individual to
an organization. the old url had redirect to the new one.
Change-Id: I19fc1e2ad30dcd97cad232c9ad58f53a523616b4
There are some comment errors, it's modify 'Captial' to 'Capital' in
keystone file, and modify 'possition' to 'position' in openrc file, and
modify 'comming' to 'coming' in stack file, and
modify 'prefered' to 'preferred' in stackrc file.
Change-Id: I0fdd539cbfff842a4ba7fca9100b881443300f9a
In the original change I said "for infra nodes, it shouldn't do
anything anyway ...". Well that was pre-Fedora 26 :)
It seems that dnf > 2.0 now intentionally throws an error when trying
to explicitly install an ignored package. Thus, as described in the
comment, take a simpler approach of skipping this on infra nodes.
pip-and-virtualenv in dib should have installed the latest pip,
virtualenv and setuptools, so we don't want to fiddle with that
anyway.
[1] https://review.openstack.org/#/c/338998/
Change-Id: Ib300b58377a0d0fe1bd7444c71acdb9a87dc033b
After looking at these for I9881f2e7d51fdd9fc0f7fb3e37179aa53171b531 I
found them not as useful as they could be.
Fix the CustomLog command, that wants the logfile then the format
string (or a nickname, which the LogFormat line wasn't setting). Use
standard micro-second timestamps, and trim the access log to have more
relevant info.
Change-Id: I9f4c8ef38ab9e08aeced7b309d4a5276de07af4b
In trying to debug periodic gate instability of CentOS, I noticed that
it is using the prefork mpm, while Ubuntu is defaulting to the
multi-threaded worker mpm.
One of the problems seems related to 502 proxy errors from the TLS
proxy. We see out-of-sync timestamps in the centos TLS proxy access
logs, which might be innocent behaviour based on the prefork model or
indicate something else.
Before going too deep down this rabbit-hole, I think it is better for
consistency to use the same mpm model on all our platforms, and start
debugging from there.
Change-Id: I9881f2e7d51fdd9fc0f7fb3e37179aa53171b531
If a project manually configures the oslo.messaging transport url for
notifications it should use 'get_notification_url', not
'get_transport_url'. get_transport_url should only be used to obtain
the RPC transport address.
Change-Id: I77772dfa9f30a3db2db6d0387260dfe3452a26ef
Closes-Bug: #1708754
This should now be able to be discovered from the service catalog,
there is no reason to set it.
Change-Id: I7383b589fbcef9423beeab735db42c594f7b56fd
get auth from context for glance endpoints:
Depends-On: I4e755b9c66ec8bc3af0393e81cffd91c56064717
When using resource classes to schedule baremetal nodes the baremetal
filters like ExactRam etc should not be used. This patch disables them
in the nova config if devstack is configured to enable ironic resource
classes.
Change-Id: Ic262ccaf8b541308042d61113a953653d2261964
With cell v2, on initial bring up, discover hosts can't run unless all
the compute nodes have checked in. The documentation says that you
should run ``nova service-list --binary nova-compute`` and see all
your hosts before running discover hosts. This isn't really viable in
a multinode devstack because of how things are brought up in parts.
We can however know that stack.sh will not complete before the compute
node is up by waiting for the compute node to check in before
completing. This happens quite late in the stack.sh run, so shouldn't
add any extra time in most runs.
Cells v1 and Xenserver don't use real hostnames in the service table
(they encode complex data that is hostname like to provide more
topology information than just hostnames). They are exempted from this
check.
Related-Bug: #1708039
Change-Id: I32eb59b9d6c225a3e93992be3a3b9f4b251d7189
When configured for superconductor mode, which is the default,
nova-compute can't reach the MQ for nova-scheduler so there is
no point in even enabling the track_instance_changes code since
it's a waste of time as the scheduler will never get the message.
Change-Id: I2662ebd47323428b403d3c2236bec78f1fb1050f
Currently, stack_install_service will accept any service name. This is
problematic because a project plugin can pass an invalid name without
noticing. This has been the case in ironic-inspector[0].
This commit ensures that stack_install_service will not silently fail
when passing an invalid service name.
[0] https://review.openstack.org/#/c/424680/
Change-Id: I1a8105bdbaf4aecb630df08da416808bf7180824
Closes-Bug: #1659042
This patch fixes path to mlock_report.py. Also add python-psutil
to dstat depends as it is required by mlock_report.
Change-Id: Ia2b507a7b923f1e3393a9cb7746c66d39d6abfde
This started to fail due to a missing || : / --force option now
that the dependencies got fixed:
2017-07-30 19:38:37.260 | ++ tools/install_prereqs.sh:source:97 : sudo rpm -e --nodeps python-cffi python-cryptography python-pyOpenSSL
2017-07-30 19:38:37.293 | error: package python-cffi is not installed
2017-07-30 19:38:37.293 | error: package python-cryptography is not installed
2017-07-30 19:38:37.293 | error: package python-pyOpenSSL is not installed
Change-Id: Ia59afb7ee564cf2044ebdb3c5ad3e54ee91d1222
The actual logic of launching a singleconductor didn't get all the way
to the launch of the conductor itself, so we were still launching 2
conductors in the Ironic case. This attempts to fix that.
Change-Id: I7ddb123dbdf3e1ec9a991e474a9990d2ccbc30d3
Some environments, like grenade and ironic, need a way to revert to
the non fleet version of the conductor setup. This really comes down
to a global topology for CELLSV2_SETUP. The prefered is with a
superconductor, but allow a downgrade to singleconductor.
Depends-On: I5390ec14c41da0237c898852935aba3569e7acae
Change-Id: I10fb048ef2175909019461e585d117b4284448c6
With cell v2, on initial bring up, discover hosts can't run unless all
the compute nodes have checked in. The documentation says that you
should run ``nova service-list --binary nova-compute`` and see all
your hosts before running discover hosts. This isn't really viable in
a multinode devstack because of how things are brought up in parts.
We can however know that stack.sh will not complete before the compute
node is up by waiting for the compute node to check in before moving
forward. This puts a few more seconds into the run, but ensures
everything is solid in multinode environments.
Change-Id: I667e6a9be3fee8bb5bfd73426eef567489e3d88d
When doing a multi-node devstack deployment starting in Ocata
the child compute nodes must be discovered and mapped to the
single nova cell (cell1). In the upstream CI we do this discovery
in devstack-gate after the subnodes are stacked, but for anyone
doing this manually we need to provide some notes on what needs
to happen after child compute nodes are stacked for a multinode
environment.
Change-Id: I68418bcf28d86c60fe42537186d89458fa778bda
Closes-Bug: #1688397
This update resolves multiple issues with python-cryptography
causing keystone server and nova deployment to fail.
This is a temporary workaround until I196f025dbf1a9ac297946b8165620676645f7210
has landed and the extraneous dependency on python-cryptography (the
package) has been removed.
Change-Id: Ifb29b9089197c0429a5fc1cd08a25d2095d481f1
The only mentionable diff is the kvm alias
does not exists so we will install
qemu-kvm as with rhel7 which also exists
in the older supported fedoras.
kvm also just an alias in suse so
switching to qemu-kvm in suse as well.
Change-Id: I5c79ad1ef0b11dba30c931a59786f9eb7e7f8587
Some URLs are broken, so fix them.
The others are redirect to new URLs,
so replace them with new ones.
The config options of nova serial console proxy
have been gathered in nova/conf/serial_console.py.
So the description in doc/source/guides/nova.rst
is fixed.
Change-Id: Ifd81cc09969341fbf8f135a913fc6003b94e0acc
Otherwise, we've seen intermittent "Unable to establish connection"
failures, with the main devstack log reporting things like
2017-07-19 13:54:29.973 -> start proxy service
2017-07-19 13:54:30.082 -> start OSC to store temp url key
2017-07-19 13:54:31.908 -> OSC reports failure
Meanwhile, the s-proxy screen session tells us things like
Jul 19 13:54:31.919988 -> start child worker
Jul 19 13:54:32.206598 -> still loading the WSGI app
... and ports aren't actually bound until *after* the app is loaded.
Add a wait_for_service call to wait for the proxy to come up.
Change-Id: I1a722de31b144797230991700e110353a2d937dd
openSUSE 42.3 is a relatively minor update over 42.2 and I'd
like to maintain it and keep it passing.
Experimental gates are being proposed, once those are passing
(which should be the case with this patch included), we could
enable it as a non-voting gate.
Change-Id: Ia421ada0ed3751c65a2a93a208e3f4a43edf8b16
With libguestfs usage for file injection now being enabled by
default as part of I568c56dbcb62ec541661364c142eff2397e3eed7
the opensuse job started to fail due to lack of guestfs images
being available.
The error in question was
NovaException: libguestfs installed but not usable (cannot
find any suitable libguestfs supermin, fixed or old-style
appliance on LIBGUESTFS_PATH (search path: /usr/lib64/guestfs)
This part is being fixed by explicitly adding the missing package
dependencies to the compute node rpm package list while the maintenance
update for Leap 42.2 is in preparation.
Change-Id: Ie76ac0a51c1ee2ad6559917825dee1c7a91a3a76
Now that mysql.qcow2 has been removed, we only have 4 images to worry
about. This fixes cache-devstack element for openstack-infra.
Change-Id: Ia06f0e0679e253a1a6614f7c38abf1f5cd13991b
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This patch adds removing of the volume group
before removing the loopback device and
the backing file when performing LVM cleanup
in unstack.sh
Without this commit:
unstack.sh removes logical volumes, removes the
loopback devices and deletes the backing file
but leaves a dangling volume group
$ ./stack.sh && ./unstack.sh
$ sudo vgs
VG #PV #LV #SN Attr VSize VFree
stack-volumes-default 1 0 0 wz--n- 10.01g 10.01g
$ sudo losetup -a
$ sudo vgremove stack-volumes-default
/dev/loop0: lseek 4096 failed: Invalid argument
vg_remove_mdas stack-volumes-default failed
With this commit:
unstack.sh removes volume groups after removing
all logical volumes but before removing
the loopback device and deleting the backing file
Partial-Bug: 1441236
Change-Id: Id9c06fa50f6cad28764f5a3396f559cac9999649
We no longer host this on tarballs.o.o, additionally it is no longer
used my trove.
Change-Id: I2034e8ebc530704d6e63a231056f92e14a8654e4
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
When the ldap service is enable on local.conf devstack ldap
plugin starts slapd service using its default config on Ubuntu
and installs ldap-utils package.
Enables domain specific drivers on Keystone and creates LDAP
domain 'Users' with a demo user.
Change-Id: I8d7aa260b01f675e4ed201ef93bfd66474f4b228
This makes us start two levels of nova-conductor processes, and one per cell.
Note that this also sets the notification transport_url to the top-level mq
so that we continue to get a unified stream of notifications.
Related-Bug: #1700496
Change-Id: I08d7da843d18b426dda8a8a231039d950a4c0ce5
Depends-On: I64b600b30f6e54db0ec9083c6c176e895c6d0cc2
Depends-On: If59453f1899e99040c554bcb9ad54c8a506adc56
In Pike, Cinder and Nova will support extending the size of
an attached volume, but it's not supported by all volume and
compute backends. Tempest will also test it but uses a config
option that we need to set based on how devstack is configured.
Depends-On: Ibace6c2f91be9753a44e5f79fd013df11654851b
Related to cinder blueprint extend-attached-volume
Related to nova blueprint nova-support-attached-volume-extend
Change-Id: I52cc2952a2938ce44c442aa3e3b69a905b2b55d5
As a follow on to I4c269a7f3d63ee9a976e7c3636fc3e5e8dab9ae3; the
quoting gets tricky when putting arbitrary command-substitution
strings into saved echo-able strings. As they say, "the only winning
move is not to play" :)
An alternative proposal is to not write this into a script but just
dump info into a file. To my mind, this has several advantages --
avoid getting involved in quoting, not dropping a script into the
global environment -- it's just as easy to "cat" -- and the plain-text
file can be collected as an artifact during log collection (also moved
git commit line to separate line for easier parsing during log search,
etc).
Change-Id: Ic7391dd087657c0daf74046e4a052c53f4eb6e1a
This commit increases the socket timeout value from 4 secs to a much
higher 30 secs. This is just for sanity, the load is high when we're
seeing the wsgi.input timeouts, so uwsgi might be just closing the
socket waiting for data over the wire. 30 seconds is overly conservative
just so we can rule this out. This will likely be shrunk to a more
reasonable value in the future.
Change-Id: Iae85d3a084fb33b2a63550d6e353413e98c0b39c
Partial-Bug: #1701088
Previously the local uwsgi server mode was using uwsgi in http mode.
This was unessecary and actually not recommend by the uwsgi docs [1][2]
This is because http mode starts a frontend http process that forwards
requests to the workers running the python code. This is done for the
largely the same reasons we're using apache as a proxy and is
unnecessary. http-socket mode doesn't do this and just exposes the
workers as an http interface to the proxy. (in our case apache)
[1] http://uwsgi-docs.readthedocs.io/en/latest/HTTP.html#http-sockets
[2] http://uwsgi-docs.readthedocs.io/en/latest/ThingsToKnow.html
Change-Id: I5671687c8083fa4bdee066c07b083a0f00be532b
Per the Pike goal, switching the Cinder API control plane to
use WSGI in Apache.
Co-Authored-By: Ivan Kolodyazhny <e0ne@e0ne.info>
Depends-On: Ie8a0eeab1bf31887d6f37cf155b2d161ddfb172d
Depends-On: I14b68f36e7fcc5341bbdbcf165274d9d50f7dd04
Change-Id: I8cef6c98f9242cc38d66de0ac499490e2a237887
According to the uwsgi docs [1] for http keepalive there is a separate
option for http keep alive, and just setting connection close isn't
enough. This commit makes sure we disable http keepalive. This will
hopefully fix the random connection issues we get on image uploads to
glance, which uses uwsgi http mode.
[1] http://uwsgi-docs.readthedocs.io/en/latest/HTTP.html#http-keep-alive
Change-Id: Ic5f83c5c93f28b2bd62ca9ac96ca8c87797ea5c9
Closes-Bug: #1701088
Unless NOVA_USE_MOD_WSGI is False, run nova-api and nova-metadata
using uwsgi.
Because the metadata server is always expected to run on a port and
without a prefix, we have it configured to use uwsgi but not to
proxy from apache: uwsgi listens on the configured port itself.
uwsgi process that listen themselve do not need a socket or to
chmod-socket, so those config lines have been moved to the block
that is also writing proxy configuration for apache.
Because this change only uses uwsgi for nova-api and nova-api-meta,
nova-api-meta is set to default to enabled in stackrc because the
nova-api wsgi application used by wsgi only presents the one service
(osapi_compute).
If NOVA_USE_MOD_WSGI is False and tls_proxy service is enabled,
nova-api is run on an internal port reached via the tls_proxy.
Depends-On: I8ff08d61520ccf04e32dcd02f4cecc39dae823cb
Change-Id: If2d7e363a6541854f2e30c03171bef7a41aff745
nova's instances directory may be a shared directory (ie nfs), in
these cases, we do not want to call nova_cleanup at startup since it
deletes everything under $NOVA_INSTANCES_PATH.
The nova_cleanup routine will still be called by the clean.sh script
which is fine since we're presumably cleaning up the whole openstack
cluster at that point.
Change-Id: Ieb4e5d0508d4ed4c5349c497554c5da2993c9cb0
Closes-Bug: #1649389
This reverts commit 15b0a5f1eb.
The change had an assumption that a service
is properly configured even when it isn't enabled.
The assumption is not true.
Change-Id: Ib5a8ffe63eaec15bc29bfdd133db7169507bab82
Closes-Bug: #1698129
In Fedora mariadb, cracklib has been enabled [0] in order to verify the
password strength.
Disable cracklib in Fedora devstack in order to allow simple passwords
in dev environments.
[0] https://src.fedoraproject.org/cgit/rpms/mariadb.git/
commit: 9442da192282aa74f43e86c96202109a173bbaba
Change-Id: I2d5e965f0f19f86992794eec78134e862899c931
We run glance behind uwsgi. This means that the URL glance knows about
itself is wrong, and version discovery fails. Set the public endpoint to
the value of GLANCE_URL which should always be correct.
Change-Id: Ia7c69024a0ef6cc0fdc284ffcd06eee5678a1007
There is confusion about where installation of new libraries should
end up, to prevent lots of little files being added make a
lib/libraries which is the old lib/oslo. Put compat functions and
includes in place to help with transition.
Change-Id: Ieeab605d187ef6aec571211ab235ea67fa95a607
This separates out the install phase early from the start phase to
make this mirror other services in devstack.
Depends-On: I4124dc7e3fd3b4d973979da85209ec991c0f8c4b
Change-Id: I76f8740448b25a48869ee80006e826baa6cafc2b
In Multi host deployments, it is possible to run ETCD in a different
host than the SERVICE_HOST (where all the controllers run). This patch
brings that distinction.
Change-Id: I15fe6f25eedf1efebaab81cce26b080577b856cc
Signed-off-by: Antoni Segura Puimedon <antonisp@celebdor.com>
The OSC number remain high, and it's useful to understand how much
time we spend making OSC calls, especially to surface it relative to
other items. The way we embed this in our code makes it hard to
instrument.
This patch creates a wrapper function for OSC which collects the timings
then aliases `openstack` to that function. This means any invocations of
the openstack utility goes through our function while devstack is
running. Because this is an alias it only affects the stack.sh shell and
any subshells.
This also moves the time tracking infrastructure to count in ms,
instead of s, because some of these operations are close enough to a
second that rounding early is losing way to many significant
digits. We divide by 1000 before reporting to the user.
Change-Id: Ic5f1844ce732d447ee980b3c9fdc417f72482609
In the helper functions to check if roles are set and if not add the
role and return the id we weren't actually checking if the role was set.
The reason for this was we grepped for name values while outputing only
uuid values with OSC. Fix for this is straightforward, we just add the
--role argument to OSC which will filter for us then we don't have to
use a grep on the wrong value type.
Change-Id: I2691b347d2a6273100deb4a1750ab353a8e49673
We are trying to keep better track of what pieces of devstack consume
the most time. Add the db sync commands to the time tracking as they run
the database migrations which can take more time than expected.
Change-Id: Ib92f2b8304ccf703712d45fd7207444de3599e2d
Because C.utf8 is not everywhere and is sometimes called C.UTF-8 (just
to confuse people) use en_US.utf8 which is in most places. This isn't
language/region agnostic but gives a consistent unicode aware locale to
devstack.
Change-Id: I67a8c77a5041e9cee740adf0e02fdc9b183c5bc4
fixes-bug: 1697733
Commit 5edae54 introduced the usage of systemd in Devstack. This allowed
the transition away from 'screen'. Systemd needs "user unit files" to
describe the services. Currently, those unit files get only created when
an openstack service (n-cpu, c-sch, g-api, ...) is in the list of enabled
services (`ENABLED_SERVICES`). This means, when Devstack is fully stacked,
there is no way to start the systemd unit of an openstack service which
is *not* in that list.
This commit changes that behavior, and creates the systemd unit files
independently of the list ENABLED_SERVICES. This means, when Devstack
is fully stacked, I can start a systemd unit of an openstack service which
wasn't in the ENABLED_SERVICES list. This allows more flexible lifecycle
management of openstack services in the gate, which is useful for tests
which test components which are not in the "default configuration" (e.g.
the "nova-serialproxy" service).
The `clean.sh` script purges all traces of systemd user unit files created
by devstack.
Change-Id: I0f7e1ee8723f4de47cbc56b727182f90a2b32bfb
Reduce bcrypt hashing rounds from 12 to 4 (minimal possilbe).
This is going to imporve a lot of perforamcne of OpenStack.
Bcrypt is hashing algorithm that is designed to use a lot of resources and
in that way stops brutforce attacks. It's exponential algorithm that depends
on amount of rounds. By default they use 12 rounds which is quite high value,
good enough for real secure production enviorments.
In case of DevStack it's going to slow down all authentication by many times.
Rally shows about 5 times slownest (adding 2-5 seconds to every authenticate)
DevStack is meant for developemnt & CI so performance is way more important than
security.
Change-Id: Id8c763d63cb91f37a774f9400f35c309f37d6f12
To reduce the total number of invocations necessary for pip which isn't
the quickest thing ever (due to needing to evaluate constraints and deps
lists and what is currently installed) combine the main installation of
software with its test-requirements.txt file which should roughly halve
our pip invocations.
Change-Id: Ibcc3264136e66d34a879ad1c90a62e1bb6a84243
Transient failures were being reported because the current lockout
period for users was too short. While this does increase the
run time IdentityV3UsersTest.test_user_account_lockout, it
allows for more flexibility if there is network latency or some
other factor that cause the lockout to expired before the
next authentication.
Change-Id: I61bc39bbc35ac414b4a72929a90845956c99eb1a
Closes-Bug: 1693917
When installing OpenStack via DevStack on XenServer, we need to
some preparation operations in dom0 which will refer the function
in devstack/tools/xen/functions file, but we are planning to move
the whole folder of tools/xen from devstack to os-xenapi, so it
this patch is to moving the dom0 related operation to os-xenapi
repo first.
Change-Id: Ib59d802a7a4eab4ccce0e29d80f29efa4655bc0b
Depends-On: I712ee74ce945859ba5118e09b7d9436ca2686cb7
The old implementation for is_$service_enabled simply checked if any of
the subservices were enabled and if so the service was considered to be
enabled. This makes disabling services complicated as it means you have
to list every single subservice which can and do change over time.
Instead also check if the generic service name is in the disabled
services list and if so don't treat the service as enabled.
Change-Id: I7fe4dfca2cd9c15069d50a04161a29c5638291cb
The scheduler_driver option has been moved and deprecated. This
change uses the new group and name for the option.
Change-Id: I27aeff5911510c9f47191acaa0c0b5b71f977cd7
We required initially 42.2 test updates to be enabled as the
liberasurecode-devel update wasn't released. It is now released
so we can stop pulling that part in.
Change-Id: I4e514e317da8a95809593a49c6dce619bc4c021f
As part of getting swift's functional testing to work properly through
the tls-proxy we need to increase the allowed request header size in
apache. This was a non issue without tls proxy as requests hit the
eventlet webserver directly which was configured via the swift config
which sets this relatively large limit (by default devstack configures
swift to have a header size limit of 16384).
Now we pass in an optional parameter to start_tls_proxy that includes
the desired header size. lib/swift then passes in the value it also
configures in its swift.conf.
If not explicitly set we default to 8190 which is apache2's default.
Change-Id: Ib2811c8d3cbb49cf94b70294788526b15a798edd
openSUSE 42.2 passes testing on the experimental gate and
in order to add it as continuosly tested target we need to
add it to the positive list of tested distributions.
Change-Id: I46f94cfad828534f324994c3d21bddff40e8f9a2
Because the swift functests (which use test.conf) run out of a
virtualenv they don't get access to the system wide trust of the
devstack CA. Handle this by explicitly configuring the cafile to trust
in the test.conf file.
We also set the web_front_end to apache2 as that is what is terminating
TLS for us. The tests handle different web server behaviors using this
flag.
Swift's functests will need to read these values in and properly
configure things on its end.
Change-Id: I4cdba36ccab6acd76205184882ee29e4f1e12333
Otherwise those services, notably n-cpu, will try to register
resource providers before placement is ready.
Change-Id: I89fd4fa42baf3d19ee209c59cd85b97adb97c58b
Closes-Bug: #1695634
The scheduler_default_filters config option moved out of the
DEFAULT option group into a more specific group, and the old
option is deprecated as a result so we need to update our usage.
Change-Id: I5d6574d19c3f16abadddb19f34cb645dcdcc07f4
Function "_install_etcd" is trying to use "files" directory
to download a file. Instead of this, this directory should be
$FILES, which is defined previously in parent script.
TrivialFix
Change-Id: I643ce3b9aba1f65f03524430c748bf120d071509
volume_clear is currently set in the DEFAULT section,
but this is a backend specific setting, and therefore
needs to be set in the backend config section.
Change-Id: Ifa3a659bb4768b8915a0f23e7f14b0f3348d93d2
Previously we didn't block out the c-api startup code because the
devstack functions to start services check that for us. However, since
the cinder devstack code checks the service is up and runs the tls proxy
if tls is enabled we need to block it all off to avoid doing those
things if c-api is disabled.
Change-Id: I1c4f22f785af96caaf4baa21ff28714b9afd3458
The upstream CI runs exclusively on nodes with x86 architectures, but
OpenStack supports even more platforms. One of them is the KVM
on s390x (IBM z systems), which is supported since the *Kilo* release.
This change describes the additional settings in the ``local.conf`` file
to enable Devstack on that platform. This is useful for PoCs.
Change-Id: I943b552ca2e36210ac57f36c16db930eb5e58623
This adds packages to suse for systemd python linkages as well as
apache2 and which. And configures mod_proxy and mod_proxy_uwsgi with
a2enmod.
We also properly query if apache mods are enabled to avoid running
into systemd service restart limits. Enable mod_version across the board
as we use it and it may not be enabled by default (like in SUSE).
Also in addition to enabling mod_ssl we enable the SSL flag so that TLS
will work...
Finally we tell the system to trust the devstack CA.
Change-Id: I3442cebfb2e7c2550733eb95a12fab42e1229ce7
This commit adds support for deploying glance as a wsgi script under
uwsgi. To get around limitations in the uwsgi protocol when using
python3 for chunked encoding we have to setup uwsgi in http mode on a
random port listening on localhost and use mod_proxy to forward the
incoming requests. The alternative approach of having apache buffer the
requests locally with the send_cl option with mod_proxy_uwsgi only
worked on python2 and also has the limitation that apache is buffering
the entire chunked object, which could be several gigabytes in size.
Depends-On: I089a22a4be4227a551c32442dba27c426f54c87d
Change-Id: Ie98fb7da5e8ecfa49cd680b88139cb7034d5f88f
This commit adds a new function get_random_port to return a randomly
available port from the local port range.
Change-Id: Icaed180cc14602a74cdb3fd3456b690d8a4c729c
We have to do silly overrides of cert locations for requests for
reasons. If we are running under python3 then we were previously looking
in the wrong location for the requests certs. Update the cert fixing
function to properly use python3 to find the certs if python3 is
enabled.
Change-Id: Id1369da0d812edcf9b1204e9c567f8bfe77c48b2
The configuration guide link has syntax problem in README.rst,
and the click the link will lead to page 404.
Fix the syntax problem
Change-Id: I47a1641a6898930dca508cdac98b1b43c05dc446
On ubuntu contents of /var/run do not persist between reboots. Devstack
uses /var/run/uwsgi as home for wsgi sockets. This means that after
rebooting the machine services, that rely on uwsgi would fail to start.
Currently it affects keystone.service and placement-api.service.
This patch changes delegates directory creation to systemd-tmpfiles,
which would run on startup.
Change-Id: I27d168cea93698739ef08ac76c828695a49176c7
Closes-Bug: #1692767
With cinder supporting this now, start logging global_request_id in
systemd logs. It will be None for all the services until the work
starts coming together, but it is safe to do.
Change-Id: Ic6ba1a42da88c03e43d89658b453f6a0b353e0db
The trove development environment is typically a linux VM within which
openstack is installed, and trove launches guest vm's within that
environment. To make it possible for these vm's to launch in one human
lifetime, one must set vt/x and enable nested hypervisors to use with
kvm; qemu emulation will take way too long.
The new libvirtd (v2.5.0) in Ubuntu Cloud Archive doesn't handle
nested hypervisors well and if you use it, you end up with a guest
hanging on the GRUB line.
To enable that use-case, we provide ENABLE_UBUNTU_CLOUD_ARCHIVE which
the trove developer can set (to False) before running devstack.
Change-Id: Ia0265c67bb7d2a438575a03c0ddbf2d9c53266ed
Closes-Bug: #1689370
truorfalse function from common-functions accepts default as the first
parameter. The arguments for USE_JOURNAL were mixed up and this commit
restores correct order.
Change-Id: Id3621b0e1910a625d6cfb8e81bd27bea82543ae9
In Ibbb430fb1dbf66942168e0cb52d990ab6a2eb8d7, we are adding
etcd3 as a new base service. We should drop zookeeper
and use etcd3 as the backend.
Since cinder is the first service for this tooz+etcd3 DLM
scenario and cinder uses eventlet we have cannnot use the
grpc based driver in tooz. So new CINDER_COORDINATION_URL
that defaults to the etcd3's grpc HTTP gateway based
tooz backend.
We need to hold this change until the tooz change (see
Depends-On) is available in a tooz release.
Depends-On: I6184ed193482dad9643ccb2b97133d4957485408
Change-Id: Ia187e1a86413edf25b909b6bb57e84fb4930a696
Etcd3 was enabled recently as new service in devstack [1]. But there's
no way to disable etcd3. This is required on architectures where no etcd
binaries are available (e.g. s390x). The long term goal of course should
be to have those binaries available. The short term circumvention is to
allow disabling the service in local.conf:
disable_service etcd3
[1] https://github.com/openstack-dev/devstack/commit/546656fc0543ec2bc5b422fd9eee17f1b8122758
Change-Id: I6184ed193482dad9643ccb2b97133d4957485408
Partial-Bug: #1693192
gpg verification requires network connectivity which is non
mirrorable. We try to avoid that in devstack whenever possible. A
sha256sum is a totally reasonable way of knowing if the downloaded
package is valid.
Closes-Bug: #1693092
Change-Id: Id496ab53f76444f08dc6961f1ecd25f450cc96d7
This change ensures that when uploading vhdx images, we use the
proper format.
At the moment, vhdx images are uploaded as vhd, which can be
troublesome: first because this is misleading, second because the
actual image format may be checked, having the image rejected.
Change-Id: I9578be41ea9dc252404b7553679ac527e08a0ff6
We need to handle this better, for now, just don't install
etcd in the sub nodes. We need to setup the proper clustering
mechanism if we want to have etcd3 running in multiple nodes
Change-Id: I8dd385e3c993942473e67d04367cdf74495dbeef
In master branch of noVNC project file vnc_auto.html was renamed to
vnc_lite.html Because of that nova-novncproxy looks for file that
actually doesn't exist.
We need to change branch of noVNC to latest stable, because other
projects are not ready yet to rename the path. Those projects
depends on noVNC package installed in system, but it is too old
for now for both CentOS (version 0.5) and Ubuntu (version 0.4).
The only way to make noVNC console working on Devstack is to change
the branch to stable one.
Unit test also has to be modified in order to ignore novnc repo
from checking against cloning non-master branch.
Change-Id: Iaf4761aedf93bc6b914a6a0c5cf1cfedcc29583c
Closes-bug: #1692513
ETCD_DOWNLOAD_URL is set to github url, in our CI, we can point
ETCD_DOWNLOAD_URL to a url in tarballs.openstack.org possibly
in devstack-gate
Download the etcd binaries and drop them into /opt/stack/bin and
use it from there. Cache the tgz for subsequent use (local workflow)
daemon-reload is called twice once from inside the write_user_unit_file
and then when we adjust the entries with additional things recommended
by the etcd team. We need a better way to do this in the future.
Added a TODO to verify the downloaded artifact later. The etcd team
posts gpg signature, we could verify that or run sha256sum and hard
code that in lib/etcd3 file. We would have to update it whenever we
bump the etcd3 version.
We use the public key "CoreOS Application Signing Key <security@coreos.com>"
with ID FC8A365E to verify the integrity of the downloaded file
Any jobs that need to be run on architectures where v3.1.7 is not available
should rey the v3.2.0-rcX release candidates. We can switch to v3.2.0
when it gets released.
Initial version of this code was borrowed from the dragonflow
repo:
http://git.openstack.org/cgit/openstack/dragonflow/tree/devstack
Change-Id: Ibbb430fb1dbf66942168e0cb52d990ab6a2eb8d7
As a followup to I2c78a0c6599b92040146cf9f0042cff8fd2509c3, the nova
cert service should be removed from devstack.
Without this fix, stacking will fail is USE_SCREEN=True
Change-Id: I115580352fa380b896bae290f9a4efbfe4ff0dfd
Update sphinx to the version used to build the documentation elsewhere
and turn on the option to treat warnings as errors to ensure that no
poorly constructed rst is introduced. Cap sphinx<1.6.1, since that
version has a conflict with pbr right now.
Change-Id: I19b3332229e2094988cbf8968c42a0323194a209
Signed-off-by: Doug Hellmann <doug@doughellmann.com>
This enables tcp stats (listen, established, syn, time_wait, close) in
dstat to allow us to get a high-level view of performance changes in
the system during gate runs.
Change-Id: Ifbffbed22446e7e6a3b825c18266b63d2f2e7718
DEVSTACK_CINDER_SECURE_DELETE is deprecated from liberty release.
This should have been removed after kilo-eol
Change-Id: I82c15a19f8fe0326d4a5c2a076baa6d3e53fcf32
One of the pending issues with the conversion to systemd was the loss of
log coloring. It turns out that journalctl by default strips out
characters it considers "unprintable" - including the color codes
emitted by the old-style logging. However, journalctl can be made to
print them by adding the `-a` flag.
This change makes devstack's log formatter conf settings include color
codes like the old screen-based setup used to
We also remove stackrc's setting of JOURNALCTL_F, whose usage was
removed via I6af6d1857effaf662a9d72bd394864934eacbe70.
Change-Id: I2401e267913a24d18dae355aa933072dbbdab1d8
'pip uninstall' will hang running stack.sh if it has to
prompt the user for input, use -y.
Change-Id: Ic94639e444b87fd3538463d5a51c01a0208a2ab2
Closes-bug: #1691172
This sets our default ulimit NOFILE to 2048, which is double what we
set things like mysql'd max_connections to.
Change-Id: I5126bed1e6b9f8c64db00eae4151ac61e47b1bf8
Add a table of contents help readers find the information they need
without having to read all of the prose on every page. Remove the
site-map file, which doesn't appear to be linked anywhere.
Change-Id: Ib5761c9cfdd5a083df562413d727cb4ac7547c9e
Signed-off-by: Doug Hellmann <doug@doughellmann.com>
At the Boston 2017 Summit I had mentioned that the pattern of using non
voting/experimental jobs was not working for getting new features into
Devstack. It is slow and leads people to being too conservative when it
comes to pushing new things in. Instead I suggested that since Devstack
changes are self testing we add the features, have change that enables
the feature, and if that changes passes we move forward with merging
(assuming code review is fine and necessary communication is done).
Document this process in the HACKING file so that we have something we
can point to when people want to add a new experimental job for every
new little thing (ipv6, tls, systemd, etc).
Change-Id: I5190cc3d3de4e81d52748347306133b5034d5531
When ironic-inspector is enabled through devstack, these two files
are created which needs to be masked.
Change-Id: I7a3db6fd6197da20cca1e938727d54195957ac18
Closes-Bug: #1690105
libvirt-python compiles against the currently installed libvirt. If
you upgrade that, it needs to rebuild, however it won't change
versions, so pip install just noops. Force an uninstall / reinstall of
it every time to handle potential upgrades of libvirt.
Change-Id: If34541b34aa6d55eedaf6c603fd1fe92eb887308
This option is a duplicate of public_endpoint for Cinder.
Change-Id: I8aee1b9f93a09d2e92bde80c0e413e1540723bac
Depends-On: I2a74af7906d14cbc49b8cf0a88c344ca30fcbd26
Now that we aren't using native journal messages by default, the
syslog identifier of all the uwsgi processes is the same. We should be
more explicit with those.
Change-Id: Id5406d02407b022d4016517c2e18890973876d88
There was detection code in clean.sh to only run it if a screen
session was found, but in systemd world, that's obviously not
true. This was causing me (and others) substantial confusion.
Change-Id: I204e94cd86b8c67012aabfca74796e593151c3a4
systemd-journald has rate limiting built in, but that's not sufficient
for the level of logging of OpenStack services during test
runs. Disable the rate limiting so that no log messages are lost.
Change-Id: I64599aba74c5a39276bb8f946cd236600b9cc81b
There are a couple of issues that have ended up being hit by devstack
plugin authors which we can detect and error in a much nicer way
instead of them having a cryptic systemd failure.
Change-Id: I45e4ac363aeefb4503015f9e1b57c58f79b58f40
Cinder use 'resource_filters.json' to config allowed filters,
copy the new added json file when set config files.
Change-Id: I397cb5859e2b3349af3cb07ee02b6463c6eccc35
Depends-On: 27aeba0b5d3cf64286125937e8336ba1d3b26b16
use_journal uses the systemd native path for logging, however there
are concerns that this might be negatively interacting with
eventlet. To be on the safe side fall back to stdout.
This introduces a USE_JOURNAL option which will let folks turn this
back on for testing.
This also adjusts the debug lines. When using the journal the pid
reported by systemd is correct. When using stdout, it will be the
parent process id, so we need to keep it to see which child each thing
is coming from.
Change-Id: Id7891c532bf99c099252e82d511a37a49506fea9
This removes some remnant heat code and references now that heat is
running in a plugin. Before merging this patch the heat team should
verify they got everything they needed into their heat plugin, as
there were more parts left than I was expecting.
Change-Id: I477e3a6e75591aa8ff836c28f7ef56aa1b5f8727
Everything else in our documentation is RST, we should convert the
Readme to RST as well (as github supports this, so it's fine to read
it there)
Change-Id: If2aabf629affc09b5daa570f4ca3bdf268cb53b0
Both keystone and neutron didn't yet have systemd awareness for
setting up logging (i.e. drop the extra date / time stamps)
Change-Id: Ib442c603c9afb679676976c37c2c6122201ae846
This moves the developer use case over to systemd, and updates all the
relevant docs to discuss the systemd workflow instead of screen. It
does so by defaulting USE_SCREEN=False, so will not impact people that
set it explicitly.
Change-Id: I6d664612bc2b850eb7f56852afbc841867223ab7
When transitioning between different wsgi modes, or service modes, we
should really safely stop and cleanup things that are started in any
service mode, which makes it easier to ensure that we don't leave
things around from past runs.
Change-Id: I33acbee39e1a2da2bfd79a5dd54b84a12a778be1
We should be able to operate without the identity admin endpoint,
given that in v3 it's all the same. This floats that out there to see
if we can or not.
Change-Id: Ic233f6b43dd1e3cfdadff0f18aba4ea78825a996
auth_uri is not a keystonemiddleware option, and it's use in config
files is confusing at best. Remove it for clarity.
Change-Id: Ie3a9ab30d81809363444d5f3b41588b3889dc185
As we move to enabling glance-api to use a wsgi script that might be run
as multiple processes, there are a couple places where external
synchronization is necessary. To use this we need to set the lock_path
config option from oslo.concurrency so external locks will work.
Change-Id: I9a66a8636d12037ff9aa4fb73cc3f9b9343dd7e9
In order to start making the transition in the gate make
USE_SCREEN=False also mean USE_SYSTEMD=True. We'll never actually
declare USE_SYSTEMD=True in the gate (as that doesn't exist for stable
branches), but this will let us roll over the existing transition.
We also have to install systemd-python 234 because we are recording
exception info in the journal, and all versions before that had a bug
in processing that.
Remove the somewhat pointless screen following journalctl commands. We
really don't want or need those, and they tend to build up over time.
Depends-On: I24513f5cbac2c34cf0130bf812ff2df6ad76657c
Change-Id: I6af6d1857effaf662a9d72bd394864934eacbe70
KEYSTONE_SERVICE_API is the keystone endpoint and it is what we should
use. The admin url should DIAF - but until it does, it CERTAINLY should
not be the thing we put into clouds.yaml.
Change-Id: If8196a04f852f633e0b7548793f68c92376aa6da
Switch from sha1 to sha256 and from 1024 bits to 2048 bits. Do this
because things don't like the old inseucre sha1+1024bits combo.
Change-Id: Iae2958969aed0cd880844e19e8055c8bdc7d064d
The os-fping API was deprecated in nova in
I92064cbcb5f6414da0c9d294f912a860428af698. I can't see anything
obviously using it on codesearch.
This is only in EPEL for centos, which I'm trying to remove. But I
think less dependencies is always better than more in general hence
the removal.
This is essentially a revert of
Ibdc7479a9038321e4fc3953774a6f3e1dac90530
Change-Id: I163fc48c860bae2a92c83cfdaed26b2e54630e20
Centos/RHEL 7 doesn't support tgtd. While the packages are still in
EPEL, there's no point in testing because nobody runs like this.
Switch cinder to use lioadm which uses LIO, and update package
installations.
Depends-On: I964917d13d9415223845ac17eb804ee7faceaf6f
Change-Id: Idc5a4e856bfc93e9dc650d565a98a8e9b3df3481
It's not only using our upstream caches that you might get an old
libvirt-python wheel that is incompatible with UCA. Move the ignore
out of the mirror check to apply it globally.
This is an alternative to Iba301a8c80c9ed584f5fb5a816f3d2cf5f5f0e77
Change-Id: I588b1e8e49aa60f3ce976dc1b6c8013ba1d88079
The DEFAULT_IMAGE_NAME variable is used to reference the name of the
default image in glance after it has been uploaded by devstack. It is
used both inside and outside of devstack for that purpose. However, when
configuring tempest there are some tests which also do image uploads and
need a filename for specifying which file they should upload into glance
for testing purposes. Previously we were just using DEFAULT_IMAGE_NAME
for both purposes, but this causes a conflict if the name of the image
we upload into glance does not have a file extension. So instead of
conflating the things this commit differentiates between them and adds a
new DEFAULT_IMAGE_FILE_NAME variable to use for this purpose.
Change-Id: Icf74badcf2093d8c75db538232b10b3ac7b86eb8
If the current interface has a default gateway configured is
determined by the regex
default.+<interface-name>
If for example 'enc1' is used, but also an interface 'enc1800' is
present, the regex will also match the 'enc1800' default gateway.
This patch fixes this by looking for <interface-name><white-space>.
This way 'enc1800' is not matched.
Change-Id: Id1d58f5be6296c3a37aef788359ae8fe0fe11d8b
As described in [1], it seems that mod_wsgi is not "graceful" reload
safe. Upon re-init, it can end up in a segfault loop.
The "reload" (not *restart*) after setting up uwsgi was added with
I1d89be1f1b36f26eaf543b99bde6fdc5701474fe but not causing an issue
until uwsgi was enabled.
We do not notice in the gate, because the TLS setup ends up doing a
restart after this setup. In the period between the
write_uwsgi_config and that restart, Apache is sitting in a segfault
loop, but we never noticed because we don't try talking to it. Other
jobs that don't do any further apache configuration have started
failing, however.
Looking at the original comments around "reload_apache_server" I'm not
sure if it is still necessary. [2] shows it is not used outside these
two calls.
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1445540
[2] http://codesearch.openstack.org/?q=reload_apache_server&i=nope&files=&repos=
Closes-Bug: #1686210
Change-Id: I5234bae0595efdcd30305a32bf9c121072a3625e
This defines a new function get_notification_url, which returns the URL
of RabbitMQ when you want connect to it, and uses in
ceilometermiddleware. This fixes an issue when we try to use AMQP for
RPC, but not for notifications.
Change-Id: I14450b2440806a17a90e5ddefc243868fdbe4f2c
On Centos, apache has a private view of /tmp and thus can't see this
socket, causing keystone to fail. This happened after
I46294fb24e3c23fa19fcfd7d6c9ee8a932354702.
Move it to /var/run.
Closes-Bug: #1684360
Change-Id: I47f091656802719c259752454ec88bf50760b967
The -ssh job with ssh validation enabled has been quite stable
for a while now [0] so I think it's time to add ssh validation
to the integration gate to prevent regressions from lurking in.
Doing this in devstack ensures that the change only affects
master as we didn't test ssh validation on on stable branches.
[0] http://status.openstack.org/openstack-health/#/g/build_name/gate-tempest-dsvm-neutron-full-ssh?duration=P3M
Change-Id: I187e560911f5d5d482eb7959e5174068c4c9a801
This makes openrc more robust for the grenade scenarios by having a
sane fallback when stackrc is not found.
Change-Id: I297ba519d581d2b6fb4d80d59434acace054bada
It turns out that we ended up with duplicate versions of this function
merging on top of each other within 3 days, and gerrit didn't catch
it. Boo gerrit. Boo bash.
Change-Id: Ic6aa2f9bafdec906de2bc51d5929beeec48a6a40
Swift proxy logs to syslog during the devstack-gate tempest runs. To
better capture the swift logs increase the rsyslog buffer size to 6k
bytes allowing for longer messages like tracebacks.
This was setup by openstack-infra previous during our diskimage
builds.
I03e42964e14d9f930c07ed047851bdf775639c59
Change-Id: Iaa232335865410600c93f47d4777ed4f1bce08e2
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Until we can test with a version of swiftclient that knows how to eat
auth_uri, swift still needs a working gate.
Change-Id: I09f9ad5c87b542df962a79898e06fbf1e968b1e3
Related-Change: I46294fb24e3c23fa19fcfd7d6c9ee8a932354702
Related-Change: Ie427f3b0b9eb834ff940fa5d52444a5a6cdcab15
Things like SERVICE_PROTOCOL and KEYSTONE_AUTH_PROTOCOL shouldn't
really be exported in openrc as they encourage using them directly to
build up keystone urls instead of actually using the OS_AUTH_URL.
Remove them.
Change-Id: I4b7cc680f7f14dae29b706a227be540c9e212cad
a2dissite will return a non-zero error code if the site that is being
disabled is not currently enabled (that is, if the conf file for it does
not exist). This can happen during development if you've been messing
with files by hand. Rather than exploding out of a ./stack.sh, accept
the missing file as meaning "it's disabled" and carry one. The rpm
version of disable, which does not use a2dissite, does this already.
Change-Id: Ie5dfd42efdff4bdba5ffaa765af000dd8e1d596e
The removed TODO was talking about USE_SYSTEMD, not WSGI_MODE.
WSGI_MODE makes sense, so the TODO has been done.
Change-Id: Ib574ef123ea4c82d4d88012c990cd1ad660d7879
This makes keystone use the proxy uwsgi module when running in uwsgi
mode. It also introduces a new stackrc variable which is WSGI_MODE
that we can use to control the conditionals in services that current
work with mod_wsgi.
Also update retry timeouts on proxy pass so that workers don't disable
their connections during polling for initial activity.
Change-Id: I46294fb24e3c23fa19fcfd7d6c9ee8a932354702
This converts the placement API to use the new WSGI_MODE variable
(which is not introduced until the next changeset). We do this so that
placement and keystone patches can be reviewed independently, but
there are some hidden coupling of mod_wsgi setup which happens only in
keystone, so if we do keystone first, it breaks placement.
Change-Id: Id5b2c67701bcc7b12c8e3764c7199d10f85df80f
The uwsgi proxy version that comes with Ubuntu xenial is too old, so
we have to build it from source. This is a temporary solution until
the next LTS.
This lays the ground work for using it in keystone.
Change-Id: I00fb1759e6988c7df0ce0f3df5ff1ce9fd7cd381
When an apache worker gets a proxy error, it will not retry talking to
the backend server until the retry timeout expires. We bring up the
proxy server *before* the backend server, and poll it. If we are
running a small number of workers, there is a likely chance that we're
going to hit one that errored before the backend was up, thus failing
for now real reason.
Set this to 0 instead to mean always retry failed connections.
Change-Id: I9e584f087bd375f71ddf0c70f83205c425094a17
Ref: https://httpd.apache.org/docs/2.4/mod/mod_proxy.html#proxypass
tls-proxy is the way we're now doing a standard install using https
between services. There is a lot more work to make services directly
handle https, and having python daemons do that directly is a bit of
an anti pattern. Nothing currently tests this in project-config from
my recent grepping, so in the interest of long term maintenance,
delete it all.
Change-Id: I910df4ceab6f24f3d9c484e0433c93b06f17d6e1
Instead of this code all existing in keystone inline, factor out into
a dedicated set of functions, and make keystone use this. This drops
uwsgi supporting https directly, but that's not going to be a
supported model going forward once we get to proxy only anyway.
Change-Id: I1d89be1f1b36f26eaf543b99bde6fdc5701474fe
We're now in a systemd world where systemd is managing the restart
effectively, there is no reason to be tricksy with apache now that
we're not working around weird upstartd issues.
Change-Id: Ifadfd504eb10a90db5177ea9180b9cd8331a2948
We're going to want to start using it by default so just start with
always installing it. This should not negatively impact anything else.
Also had to fix the test using cowsay, now that cowsay depends on
cowsay-off.
Part of uwsgi in devstack.
Change-Id: I8306a992d9d006bc0130a255145a6880065aa0df
Allow cursive to be installed from git instead of pip.
The barbican-tempest-plugin, which uses cursive indirectly
through nova and glance, would benefit from the ability to
use cursive from git instead of pip.
Change-Id: Icae7d310f1ee392d080e7c8e421a26d7c0ef4727
Allow castellan to be installed from git instead of pip.
Castellan has recently been moved under the oslo framework,
and the barbican-tempest-plugin tests which use castellan
would benefit from the ability to usd castellan from git
instead of pip.
Change-Id: I96edca90c61aec84637b7b1ce842eff04c521923
The swift functional tests use a config which requires keystone ports,
we're about to make those go away. This exposes the actual auth_uri to
swift for consumption.
Change-Id: I5868dfdb8e5f0972ba04e359d212b04351502436
We recently disabled EPEL in openstack-infra, enable it again.
Change-Id: I213b302b34b740354d63b69e8ac7f4e1b3d3cdd7
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
It is possible some CI system are using an http_proxy. Use the helper
function to cover this use case.
Change-Id: Iee685147ca0244fc7de328a765f937602223de20
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Libvirt live migration requires netcat. It appears that newer UCA
packages may not automagically pull this in so explicitly list it as a
dependency of nova compute here. Note that netcat/netcat-traditional do
not appear to work and netcat-openbsd is required.
Change-Id: If2dbc53d082fea779448998ea12b821bd037a14e
Give instructions to add permissions for stack user without touching
main config.
Closes-Bug: #1680459
Closes-Bug: #1681418
Change-Id: Idd27e684e63c616466de28c07551729a1e091bdd
We have had issues with libvirt 1.3.1 which is stock on Xenial. Try
using 2.5.0 from UCA instead.
Related-Bug: 1643911
Related-Bug: 1646779
Related-Bug: 1638982
Change-Id: Ia4434541c71f050fe1ffb54f4c4c1e302391d00b
The change to remove references of XEN_INTEGRATION_BRIDGE
(If5886e3711765a97f40f20e478f958b988b5a620) unfortunately left some
code which should have been removed. This remaining code caused an
error in some situations when deploying from scratch (which the CI
avoids for expediency)
Change-Id: Ia568462c9cca8cff8fcfada8148d185609d61a7d
This is required to install python-openstacksdk from source for testing
other projects against master.
Change-Id: Iee7b043ac7d381dadf89d26098f69e935ed81d6b
These are needed for oslo.log to use journald support. They are
*probably* already installed, but just in case we force install them.
Change-Id: I0dc66bd2628ff4b3e1caa7ab4366d7f36ff7ea94
This is going to be a new option in oslo.log, which we can start
setting early to make it take effect.
Change-Id: If0e5e4717a1810c759058f33608fbac7543f2d85
uwsgi is a different service type under systemd and shouldn't be run as
a standard oneshot type. The uwsgi docs outline a good pattern for
writing systemd unit files:
http://uwsgi-docs.readthedocs.io/en/latest/Systemd.html
This commit takes those suggestions and creates a separate path for
writing uwsgi unit files.
Change-Id: I9b541b86781afdded311dba058cedd783e1a0dfa
It looks libvirt/qemu requires mode 0755 to functioning properly,
but DATA_DIR won't be set to 0755 if it is different from default.
Change-Id: I37ca0b02b6a75b3756860b547e84c37ccfc99d13
Closes-Bug: #1677421
This adds a flag and basic config for enabling coredumps for libvirt.
Partial-Bug: 1643911
Co-Authored-By: Matthew Booth <mbooth@redhat.com>
Change-Id: If7cd54e804a5a389a0d82a325b58f5b41b8ef0db
In an effort to reduce memory consumption enable KSM by default. The
biggest win here is when using libvirt with nova or ironic with its
fake baremetal instances. In theory any process that runs duplicates
with mergeable memory will benefit though.
Change-Id: I4c5addfd3e83b6516023b36cbaabd5169f0d5ceb
Currently Devstack starts all Swift services, including those
in charge of "consistency convergence" (remember Swift is eventually
consistent), data scrubbing, hard-deletion (*-reaper services)
cleanup.
But when running with Replication Factor 1 some of those services
are not needed at all. Besides, the fonctionnalities provided by
some of these services are not tested at all (neither in Tempest
nor in Swift functional tests).
Thus, in light of saving some Mo of RAM, this patch introduces a config
flag to start only a minimal set of Swift services, just what's required
to make all of our current tests pass.
The default value for this new config flag is set to start all services,
that is to maintain Devstack's current behavior.
For sake of completeness, here is the list of services that are not
going to be started is the config flag is toggled, and the associated RSS
according to our peakmem_tracker
40004 swift-object-replicator /etc/swift/object-server/1.conf
34320 swift-container-replicator /etc/swift/container-server/1.conf
33584 swift-object-auditor /etc/swift/object-server/1.conf
33328 swift-object-reconstructor /etc/swift/object-server/1.conf
31936 swift-object-updater /etc/swift/object-server/1.conf
31492 swift-account-reaper /etc/swift/account-server/1.conf
31076 swift-account-replicator /etc/swift/account-server/1.conf
29540 swift-container-updater /etc/swift/container-server/1.conf
29220 swift-account-auditor /etc/swift/account-server/1.conf
29036 swift-container-auditor /etc/swift/container-server/1.conf
So we are looking at saving at most ~350Mo of RAM (could be less
because RSS doesn't account for shared memory).
A follow-up patch will soon be proposed in devstack-gate to not run
those additional services in our Gate jobs.
Change-Id: I8a0d03ac0296a74e38efd185beb8513866eaf0c4
As of Id6e3c0ac54b21d85e68625a5b52fe2559fb70f24 keystone's policy
file is empty and it is no longer required at runtime. This commit
updates devstack to not deploy a policy file for keystone because
devstack doesn't specify any policy overrides. Instead, we can remove
the sample policy file and rely on the defaults that have been
registered in code. This is the same approach nova took with policy
in I85a251376dfe38caa4b100861bf764014a98bc37.
Change-Id: Ib1d9a51a78e2a84a3d7294dc8782605a681fa9e8
During the PTG there was a discussion that the screen developer
workflow wasn't nearly as useful as it once was. There were now too
many services to see them all on one screen, and one of the most
common service restart scenarios was not restarting one service, but a
bunch to get code to take effect.
This implements a 3rd way of running services instead of direct
forking via bash, or running under screen, which is running as systemd
units.
Logging is adjusted because it's redundant to log datetime in oslo.log
when journald has that.
Swift needed to have services launched by absolute path to work.
This is disabled by default, but with instructions on using it. The
long term intent is to make this the way to run devstack, which would
be the same between both the gate and local use.
Some changes were also needed to run_process to pass the run User
in. A hack around the keystone uwsgi launcher was done at the same
time to remove a run_process feature that only keystone uwsgi uses.
Change-Id: I836bf27c4cfdc449628aa7641fb96a5489d5d4e7
psutil is only installed under python3 for the 3.5 gate jobs. Call
mlock_report.py with $PYTHON so we support both environments.
Updates to mlock_report.py for python3 compatability
Change-Id: If7926ce6a2996b766c49b010a7f6640ae624f860
Now that we don't support Ubuntu Trusty anymore, we can remove
the ebtables race workaround.
Closes-Bug: #1675714
Change-Id: I70483f871e35fcaa933d1b7bac7dbb396aa22cef
Since Feb 25th devstack supported operating systems changed
due to Nova increasing its minimum required libvirt version.
Further details see: I6617283afd798af37e64913b7865cea3c8a62aba
This patch is to update versions on devstack documentation.
Change-Id: I12bb59b0903a728376ee9422213c2903b9138249
The package libvirt-bin is a transitional package in Debian and should
not be used anymore.
Ubuntu Xenial is an exception here.
Because of that this change also adds the possibility to use "not:" to
exclude distros in files/debs/* just as "dist:" limits distros.
Depends-On: Icc59ea79f54d4ff8751f2e353ee3530fff3d961e
Closes-Bug: #1673840
Change-Id: I3998a7178d14ec40eae5cb199d66da9546cd6ccf
In order to get memlocked pages this needs to be run as root, just
start it as root so that we don't have issues with the inability to
run sudo later in the run.
Change-Id: I7adab8cbb6d89d4717e427aec22e316d27bea075
This commit change check of 'running_in_container'
method so that other services ironic, nova and neutron
will not break.
Change-Id: I42eb587cfaebf37944cb10e459b8b8f7b4b4e4ba
Previously we use a specific integration bridge for neutron ovs agent
which is running in compute node, but this isn't necessary, this
patch is to remove the specific integration bridge for XenSever and
remove the custom integration bridge definition
Depends-On: I675565e1ea6c887d40d7a53f62968c4aa385ecca
Change-Id: If5886e3711765a97f40f20e478f958b988b5a620
single-machine.rst and index.rst
Before this, one had to create /opt/stack
manually and chown it to the stack user.
Now it is created when the user is created.
This is the same way the multi-node guide
handles it. A stack group is created too.
Change-Id: I5363d81c8fb38796f565cc6ebf6ab2dee2673989
Closes-Bug: #1673787
Closes-Bug: #1671409
With XenServer we have two neutron-openvswitch-agent(q-agt, q-domua)
For the q-domua it is specific for XenServer, this patch is to move
the specific configurations to os-xenapi which we have devstack plugin
in that repo
Depends-On: Ic816404c84f6a8899d01a77cb67fbfb421653e6b
Change-Id: I8a31c81d9475387fe4ed7030b70b26098e588771
When redoing a stack.sh over and over again, tempest venv can get out
of sync and cause issues until deleted. We should rebuild that tempest
venv on every stack.
Change-Id: I2f66bb1a7ccf9f89e11db1326d8553589e52fbf2
Supported Debian distros (codenames) are "sid", "testing",
and "jessie", but it should be "stretch" and "jessie".
"testing" is no codename and therefore should be replaced by
"stretch".
"sid" changes all the time and cannot be guaranteed to run
correctly or is at least not tested.
Change-Id: Id4b80a055452bbff69036d4dc1adeda46ce99664
Closes-Bug: #1673810
Closes-Bug: #1674416
Instead of grepping for 'lxc' in /proc/1/cgroup, use systemd's
features. This now at least also works in LXD containers.
Change-Id: I35e807c26f0b1fbba83ddbe04cfb4901a7a95cbe
We are facing memory pressure in gate testing. Apache is fairly large so
tune its connection limits down to try and squeeze out more useable
memory. THis should be fine for dev envs, also tlsproxy is not enabled
by default so we can check that this tuning works well on a subset of
jobs before making it default everywhere.
Data comparisons done with gate-tempest-dsvm-neutron-full-ubuntu-xenial
jobs.
Old: http://logs.openstack.org/37/447037/2/check/gate-tempest-dsvm-neutron-full-ubuntu-xenial/721fc6f/logs/screen-peakmem_tracker.txt.gz
PID %MEM RSS PPID TIME NLWP WCHAN COMMAND
20504 0.2 16660 19589 00:00:00 34 - /usr/sbin/apache2 -k start
20505 0.2 16600 19589 00:00:00 34 - /usr/sbin/apache2 -k start
20672 0.2 16600 19589 00:00:00 34 - /usr/sbin/apache2 -k start
20503 0.1 14388 19589 00:00:00 34 - /usr/sbin/apache2 -k start
19589 0.1 9964 1 00:00:00 1 - /usr/sbin/apache2 -k start
Total RSS: 74212
New: http://logs.openstack.org/41/446741/1/check/gate-tempest-dsvm-neutron-full-ubuntu-xenial/fa4d2e6/logs/screen-peakmem_tracker.txt.gz
PID %MEM RSS PPID TIME NLWP WCHAN COMMAND
8036 0.1 15316 8018 00:00:01 34 - /usr/sbin/apache2 -k start
8037 0.1 15228 8018 00:00:01 34 - /usr/sbin/apache2 -k start
8018 0.1 8584 1 00:00:00 1 - /usr/sbin/apache2 -k start
Total RSS: 39128
Note RSS here is in KB. Total difference is 35084KB or about
34MB. Not the biggest change, but we seem to be functional and it
almost halves the apache overhead.
Change-Id: If82fa347db140021197a215113df4ce38fb4fd17
Now Apache2 has 5 dedicated processes for Keystone Admin and 5 for
Keystone Public. As each Apache process consumes some memory and
we arbitrarly decided 5 was a good number more than 2 years ago,
maybe now (with the recent memory pressure we feel) is a good time
to reconcider.
With 5 processes our peakmem_tracker.py script reports a max RSS size
for the "wsgi:keystone-ad" and "wsgi:keystone-pu" processes of
2 (public and admin) * 5 (number of processes) * 90 Mo (RSS of each
process) = 900 Mo.
With 3 processes, the overall max RSS for Keystone is
2 * 3 * 90 = 540 Mo.
Note that this is RSS memory, but using the "smem" linux command on
my laptop, I noticed that the USS (Unique set size, i.e RSS excluding
shared memory) is around 80Mo per process. So reducing the number of
processes will actually reduce memory consumption.
Change-Id: Iba72d94aa15ecaa87c0115ad26d6bpeakmem_tracker62d5b3bea0a
The diet seems to be too strict, jobs failing with "out of sort memory". Needs more investigation before resubmitting.
This reverts commit 1e66388c5f.
Change-Id: Ic10effaaf047eb3527082baab889772c5e57fa90
It's 2017, some of our newest OpenStack developers/users may not
even know what stable/kilo is/was.
Change-Id: I00f39cc80af7e1632293bf057d95040b6bfa48e0
Was using the '-ne' integer comparison operator to compare UUID
values. This caused error messages like:
/opt/stack/new/devstack/lib/tempest: line 226: [[: dfae26ac-1780-4677-902d: value too great for base (error token is "902d")
Change it to use '!=' string comparison operator
Change-Id: Ib7c9197dd0fe58addf33b4f82beea6de64f6b10b
glare has been removed from glance already. Now error will be raised
if enable g-glare in local.conf.
Remove the glare support by glance.
Change-Id: I9a389af194dd2b8aed75d3c921293d800f8c591b
Add a missing --subnet-range argument when creating an ipv6 provider
network. Also changed SUBNET_V6_ID to IPV6_SUBNET_ID. And remove the
--ipv6-address-mode arg because it doesn't apply to subnets on routers.
Change-Id: I82796804a06e758e458606dc9eb400bcd08ad6e4
Since it's pretty common to see blogposts recommending to mount /opt/stack
remotely or editing inline the code, adding some notes about the potential
risk of a reclone that could impact weeks of work.
Change-Id: I733d40b76fb02d8edf3719533fc8202547771871
Co-Authored-By: Stephen Finucane <sfinucan@redhat.com>
Not all Cinder backends support the 'manage volume' feature. The test that
in tempest for this feature is specific to LVM and will *not* work for
other work backends regardless of them supporting the feature.
Change-Id: I055aa66738deb5ae2fb925429cec565e3901340c
When variables use the 'declare' directive, it is by default a local
variable. While other variables have global scope.
For example:
declare -A AN_ARRAY # local in scope
foo=1 # global in scope
This causes errors to occur as some of the variables will be local only
and others will be global.
Update the code, as appropriate, so that variables using the 'declare'
directive also include the '-g' flag to have them also be global. Not
every instance of a declared variable has been updated.
Closes-Bug: #1669509
Co-Authored-By: John L. Villalovos <john.l.villalovos@intel.com>
Change-Id: I2180b68fe861ad19c6d4ec0df0f9f8a528347862
lib/keystone builds KEYSTONE_SERVICE_URI so that other services
don't need to reconstruct the identity URI. Many services already
use it, but some parts were still building the identity URI from
the different parts.
This will allow changing the identity URI to include a path
(e.g., to http://<host>/identity) in 1 place rather than in
multiple places.
Change-Id: I58cbdbe591d8869807545e0815480fc3375e0479
Before the patch, we were only configuring root_helper_daemon to point
to oslo.rootwrap, but not root_helper. (The former is used for long
running commands only, while the latter is used for short lived
commands.) This made neutron agents to directly call to sudo when a
privileged process was to be executed. This failed because /etc/sudoers
was not configured to allow anything except the rootwrap call itself.
This patch simplifies rootwrap handling in the code; it also sets
root_helper to point to rootwrap; as well as configure daemon in
sudoers. While at it, we also set l2 agent to use rootwrap too.
Hopefully, it will be enough for agents to actually configure backend as
needed.
Change-Id: Ib05a6e0e024f534d7f616d41d70fb67ecf6daeaf
The oslo.messaging docs on the notification messaging driver
says that "messaging" (1.0) is a legacy format and you should
use messagingv2 unless otherwise required for that old format.
By default we should be testing with messagingv2.
Change-Id: I3031afe7551a0c8dde46e1ccfacff445fb68e122
The change makes peakmem_tracker list processes that lock memory pages
from swapping to disk. It may be helpful when debugging oom-killer job
failures in gate in case when dstat shows that swap is not fully used
when oom-killer is triggered.
The peakmem_tracker service was renamed into memory_tracker to reflect
its new broader scope.
Needed-By: I5862d92478397eac2e61b8a61ce3437b698678be
Change-Id: I1dca120448ee87930fe903fd81277b58efaefc92
Keystone v3 API is CURRENT and the v2 API is deprecated now.
So we need to change the default config of auth_version to fit
for current API status.
Depends-On: Id5e5ed9bf4f8b0f9eb376bfc7c5801f0956da1d9
Change-Id: I801e6740258ddea2a1b628a209970e0307d39d12
Nova has been supporting glance v2 since Newton and removed
support for glance v1 in Ocata:
97e7b97210139a7f7888f0d6901e499664de02a3
We should disable glance v1 by default because there are several
test paths in Tempest which don't get run when glance v1 is
available because it uses glance v1 rather than v2.
Depends-On: I54db379f6fbe859fd9f1b0cdd5b74102539ab265
Change-Id: I7f962a07317cdad917ee896d79e49ee18938d074
Change ac5fdb4c40 introduced
a problem for gnocchi CI because the deployments steps
are now:
1. create cell0
2. start nova-api (with multiple workers)
3. install ceilometer via extras
4. ceilometer calls nova-api to list servers; at this point
nova-api getes the list of cells and caches them, which
will just be cell0
5. create cell1 via simple_cell_setup which also discovers
the n-cpu node so we can schedule instances
6. gnocchi tests create and list instances and at this point it hits
an n-api worker that only has cell0 cached so it does not
find some test servers it created and fails.
The cell0 and cell1 cells should be created in the nova_api db
before starting n-api so that when we first list instances, we
store both cells in the cache that's in n-api. This deployment
order is also how the nova docs describe rolling out cells v2
but the way we were doing this devstack wasn't following that,
or accounting for when devstack plugins are loaded via extras.
This change creates the main cell1 cell earlier in the setup
before n-api is started, and then changes to just run
discover_hosts at the end after n-cpu is running (which is what
simple_cell_setup and map_cell_and_hosts would do implicitly).
Change-Id: I38eab6707340253a10159a169ae61d34784c2d28
Related-Bug: #1669473
This simplifies neutron library code and makes it less prone to breakage
in the future. So far there are no specific known issues with existing
code per se, it works, still.
Change-Id: I28f1997d226baae902dae5ca8ee6cd4fd89efe31
We propose several MySQL configuration parameter changes (with
explanations) to reduce the memory footprint of MySQL. A demonstration
of the improvement is provided in
https://etherpad.openstack.org/p/change-438668.
As Clint provided some of the descriptions that I've used, I have
listed him as a co-author (thanks Clint). Let this serve as a warning
to all that commetors may be enlisted :)
Change-Id: Icb2d6ea91d3d45a68ce99c817a746b10039479cc
Co-Authored-By: Clint 'SpamapS' Byrum <clint@fewbar.com>
Since https://review.openstack.org/#/c/438325 landed
it only works for Centos 7, but not for other
RHEL-based distributions: Virtuozzo and, probably, RHEV.
Both of above have own version for qemu-kvm package: qemu-kvm-vz and qemu-kvm-rhev,
accordingly. These packages provide "qemu-kvm", like qemu-kvm-ev,
and, when you call "yum install qemu-kvm", they replace the default OS package.
Change-Id: I46da627c0da8925064862fdc283db81591979285
This patch enabled the "allow_global_implied_dsr_disabled" feature
flag. This is a feature flag toggle for bug 1590578 which is fixed
in Newton and Ocata. This option can be removed after Mitaka is
end of life.
Change-Id: I70e3ce79ee6d9b00cc48bb178bd423d0196f6588
Related-Bug: #1590578
Because things like nova may need to create vhosts in the rpc backend,
we need to have started and created credentials before we configure
the service.
Change-Id: I01c9c5288e197fc50a8a4a032e3a32cd166eb180
openvz is not in the nova tree, and is referencing a crazy old image,
we're going to assume that if anyone is using this they can build a
devstack plugin.
drop doing anything by default because this actually requires that we
special case things like ironic in tree to *not* do anything by
default.
Change-Id: I9d33b98263c3d52a95b9983e90eb0b341fa1d363
Add python-openstackclient to the list of packages installed under
Python 3 by default, so that jobs running with Python 3 exercise the
client that way.
Change-Id: I9778a6810bb3e4850132cfc19e583d50fed23ef5
Signed-off-by: Doug Hellmann <doug@doughellmann.com>
DHCP agent should not load core plugin config file; L3 agent has no
interest in metadata agent configuration file; etc. It's a mistake to
form a single global list of configuration files and pass it into all
processes. Every process should have its own list, that may or may not
have some files in common with other processes.
The only file that is common to all neutron processes is neutron.conf,
and we could in theory keep it into the common list. But I decided at
this point it's better to be explicit about what's loaded into services.
Also the order of arguments is important, and neutron.conf should always
be the first CLI argument, which is hard to achieve by keeping
neutron.conf file in the global list.
Plugins may be interested in loading additional files into neutron
processes. For example, dragonflow needs to load /etc/neutron/dragonflow.ini
into neutron-server. But we should not necessarily load all those files
into all processes, so such extendable lists should be per process.
Besides, neutron_server_config_add_new is already available to use to
append additional configuration files for neutron-server. That's why the
patch completely kills the NEUTRON_CONFIG_ARG variable.
Depends-On: I4bd54a41a45486a5601373f9a9cce74d7686d1aa
Change-Id: Ia3c3862399bba335db5edf9ea70f850fb2638d09
Calling enable_kernel_bridge_firewall inside a
container, devstack will crash because it tries to
load a kernel module by calling 'sudo modprobe' on
net.bridge.
Change-Id: Id4718c065d5a8c507d49f38e19c2796a64221aa4
Closes-Bug: #1662194
For the latest qemu-kvm, you have to use the qemu-kvm-ev package,
which is based off the qemu-kvm-rhev package, which is explained in
[1] but you probably can't read it. The gist is, that qemu-kvm-rhev
is a later build of kvm that is incompatible with the base version
provided. qemu-kvm-rhev is only provided with the RHV (ovirt) and
RHOS (openstack) products. CentOS rebuilds this package as
qemu-kvm-ev as part of it's virtualisation SIG.
I9a972e3fde2e4e552f6fc98350820c07873c3de3 has bumped up the minimum
qemu version to 2.1.0. It seems there is a an issue (bug #1668164)
where having the qemu-system package installed gets picked up if
installed, and reports the incorrect version to nova, causing failure.
This removes the installs from files/rpms/nova as it is all being done
in function-libvirt. We only install the qemu-kvm-ev package on
centos and remove the old work-around.
[1] https://access.redhat.com/solutions/629513
[2] https://wiki.centos.org/SpecialInterestGroup/Virtualization
Change-Id: Ide91b261f35fb19d8bd7155ca016fa3b76a45ea1
Use pip_install_gr so we get the version pinned by requirements. The
depends-on is an example of where we're trying to pin to workaround
issues.
Depends-On: I9c57c08a150571c5bb62235d502839394d53a4c1
Change-Id: I780cca681b12a3e9d228dbf2fd9fa6e8ab1a82e1
This patch creates symlinks between neutron-* files tracking runtime
dependencies for services, and q-* files. Since lib/neutron is the
new recommended way to deploy neutron with devstack, this patch made
neutron-* files real files, while q-* files are just symlinks.
Change-Id: I0f361f31160d0fee40ad3e8de873edd05173e54e
This allows post-config phase to use neutron_service_plugin_class_add
so that devstack plugins can use it in the same way for both of
neutron-legacy and its "modern" counterpart, lib/neutron.
Closes-Bug: #1667037
Change-Id: I9068fd608e82e70db8d725f92269a26920efebcb
Using the same name for two different subnet pools means that one needs
to reference them by their UUID. Choosing unique names will allow us to
use the name to reference the pool later on.
At the same time simplify the command used for pool creation by
instructing OSC to only output the value that we are interested in.
Change-Id: Idedcb6328925d44cdd0f415450ec4ebbc272401d
Devstack configures keystone for auth mechanism but don't tell
keystoneauth1 library that it should use keystone too.
In simple case, this is not an issue because some application
set 'password' by default (like the openstack cli).
But applications can have no default or another default.
Change-Id: Idd1e1d2e7546fce7531175440788a8c7cb27aec1
This patch adds haproxy package to devstack as Neutron will rely on it
for serving metadata instead of the current Python implementation.
haproxy will reduce the memory footprint from ~50MB to ~1.5MB for
serving metadata. It will be spawned for every Neutron router so,
for large deployments, it will be a significant memory reduction.
Change-Id: I36a5531cacc21c0d4bb7f20d4bec6da65d04c262
The old code assumed _neutron_service_plugin_class_add presence, as well
as used a configuration file path that is not standard (under
/etc/neutron/services/ instead of /etc/neutron/metering_agent.ini).
The patch untangles metering configuration in the new library from that
old and bad code, and reimplements it inline.
This should help the effort to switch gate from lib/neutron-legacy to
lib/neutron.
Change-Id: I0d235498af4b6a70bd5dae6ea178d5aa8ba41e80
Since for the new devstack library we still rely on some functions from
ovs_base, we need to initialize them with sane default values so that
setup works as intended and as lib/neutron-legacy behaves by default for
external connectivity setup.
Change-Id: I412ed4f988b8e03a3e3a08066375b55a6e6aa3e6
The creation of the cellsv1 rpc vhost was buried in the restart function,
which makes it hard to extend. This breaks it out into a helper method
and moves the conditional logic into the nova module itself.
Change-Id: Ib0e377aabe45c27bb6ce59ca275ce73085e8b9d2
Currently placement api section will be configured on controller
only if service n-cpu is running. It breaks multi node setup.
Closes-Bug: #1667219
Change-Id: I8b0f60f253859f704bb9831d7dac8f55df353ac7
- Remove extra spaces
- Fix a missing space in the generated option string
- Fix a fatal typo
Change-Id: Ieca1c3e3c7e2ff59089ef45435e126ce7ff4f9b5
Closes-Bug: #1667073
FakeLoggingVolumeDriver will be used for functional Cinder tests to
prevent dependencies on any storage.
FakeGateDriver is based on LVM and will be used to run Tempest tests for
such features like CG's, replication, etc.
Depends-On: I383bcdb531c7d52c0fdbb6875de73f1274a92854
Change-Id: I2dc8ea416f5eb3fcc9d2e959533497e464220ff5
Upgrade xen tool to install Ubuntu Xenial; change the upstart
task to systemd task to finish the OpenStack installation
by using devstack.
Change-Id: I8129923be3c41e7f60e9d32348a5ea8e07d4845b
This ensures we only set the nova catalog when it's not the default,
instead of also putting defaults in devstack.
Change-Id: Ibb0dcb8bae2e9223db302d7b19e8fbee4ebbf0e3
I think now is a good time to stop running Cinder V1 tests. It should
save quite some Infra resources and jobs should run faster too.
Also, remove some useless variables in lib/tempest.
Change-Id: I0edf1d88c136c3b910a5773690a603eeacb50266
Sometimes when doing worlddump would see a command line like this:
sudo ip netns exec (id: ip addr
This would cause an error to be seen in console.log:
2017-02-07 00:03:03.659570 | /bin/sh: 1: Syntax error: "(" unexpected
This is caused by there sometimes being extra data returned from the
'ip netns' command [1]. For example it might look like:
qrouter-0805fd7d-c493-4fa6-82ca-1c6c9b23cd9e (id: 1)
qdhcp-bb2cc6ae-2ae8-474f-adda-a94059b872b5 (id: 0)
[1] https://lwn.net/Articles/629715/
Change-Id: Icece442023125ef55696b8d92a975d37e358b1b4
Closes-Bug: 1653969
This reverts commit e0a37cf21e.
This didn't help fixing bug #1630664. Issue seems to be between
client<--->Apache2, not between Apache2<--->eventlet
Change-Id: I092c1bbf0c5848b50fc9e491d1e9211451208a89
This commit switches devstack to use the published qcow2 cirros image
instead of the AMI version. Using AMI was mostly a historical artifact
dating pretty far back, but in the real world no one really uses AMI
images with openstack clouds. This change reflects that and also
enables tempest ro remove its deprecated config options for using AMI
as a fallback on misconfiguration (which was just there to support
devstack's defaults)
Change-Id: Id65ebae73b28da7185cb349b714b659af51ef77f
Nova is going to increase the minimum required libvirt
in Pike to 1.2.9 in change:
I9a972e3fde2e4e552f6fc98350820c07873c3de3
Based on the libvirt distro support matrix wiki [1] that
drops support for Ubuntu Trusty and Debian 7.0/Wheezy.
Trusty has libvirt 1.2.2 and Wheezy has 0.9.12 (the Wheezy
support should have been removed long ago apparently). The
7.0 removed here is for Wheezy also based on commit
b2ef890db3.
This does not undo the check for "trusty" with the
EBTABLES_RACE_FIX in lib/nova_plugins/function-libvirt
since you can still force devstack to run on Trusty if
you specify the FORCE=yes variable.
Note that RHEL 7.1 has libvirt 1.2.8 so it won't technically
work with devstack and nova + pike + libvirt, but with the
way os_RELEASE is calculated the minor version is dropped
for RHEL distros so we just get "rhel7".
Also note that this doesn't attempt to continue supporting
Trusty or Wheezy if nova is not configured to use libvirt,
simply in order to start moving forward on devstack distro
support in general and to keep some sanity and closeness
to what we test with in the CI system.
While we're in here, we also drop Fedora 23 and add
Ubuntu Zesty.
[1] https://wiki.openstack.org/wiki/LibvirtDistroSupportMatrix
Depends-On: I9a972e3fde2e4e552f6fc98350820c07873c3de3
Depends-On: If69f99bd789e646b0261e27a8a061efde32436f7
Change-Id: I6617283afd798af37e64913b7865cea3c8a62aba
The proliferation of internal/admin endpoints is mostly legacy and
based on some specific deployment patterns. These are not used by
everyone, and for the devstack case aren't really that useful. We
should simplify our service catalog down to the minimum we need for
development.
Change-Id: Ided7a65c81b3a0b56f0184847fc82e17c29a771e
Adding sudo to the example commands in the quick start section.
Also adding '-' as su argument in order to use
the stack user's env (home).
Change-Id: I23ab38104d05c3f4c8d48b55e66cf19dc4e4f90d
This new version of CirrOS is built on top of 0.3 branch, so wrt
version 0.3.4 it includes only two commits:
- Cherry-pick of the fix for https://launchpad.net/bugs/1564948
- One extra fix to make the above working on 0.3
Cherry-picked commit is http://bazaar.launchpad.net/~cirros-dev/cirros/trunk/revision/366
A Tempest test for hard reboot in some cases hits the case where
host key are empty. This triggers bugs/1564948, i.e. the ssh
daemon does not start at all, and the Tempest test fails with
"connection refused", which is misleading.
The new version of CirrOS solves this problem as it ensure host
keys are generated if missing, and the sshd deamon started.
I tested the scenario of missing host keys in Iea74c63925be17a1df894c1a2c23f5ba2793e0c6
using a private build of what then became 0.3.5.
Change-Id: I5c154ec25555e768954538fc22b4f5d5975b2deb
Force mod_proxy to immediately close a connection to the backend
after being used, and thus, disable its persistent connection and
pool for that backend.
Let's see if that helps fixing bug #1630664 (the
Connection aborted/ BadStatusLine thing).
We already have an ER query (in queries/1630664.yaml) that should show
whether this is effective.
Change-Id: I03b09f7df5c6e134ec4091a2f8dfe8ef614d1951
Since the empty value is the default for the option, and when explcitly
set in config file, it triggers a deprecation warning for the option,
avoid setting it unless we actually need to override the new default
value.
Change-Id: If423114d7a52da29b97d1fb473a955d9d69a1a3e
Those are not called by devstack anymore. This cleanup also gets rid of
code that attempts to set external_network_bridge to an empty value,
which triggers a deprecation warning for the option since it's going to
be removed in a next Neutron release.
Change-Id: I5adcbab877b4e8742522de81b1a85acfc33160d7
For the instance which has boot by uefi, we should use
virsh undefine --nvram to undefine it. Check the libvirt
version for whether it supports nvram and use new undefine
parameters since this parameters is compatible with those
instance which don't use uefi.
Closes-bug: #1612613
Change-Id: Ibca1450e965df1481e6cd6b0d597b4323d667e60
Signed-off-by: Kevin Zhao <kevin.zhao@linaro.org>
We currently use a more permisive STRICT_ALL_TABLES mode, but that's
not what modern MySQL versions default to (i.e. TRADITIONAL):
https://dev.mysql.com/doc/refman/5.7/en/sql-mode.html#sql-mode-changes
(non-Devstack deployments will most likely use TRADITIONAL as well)
Due to the fact that we default to TRADITIONAL in oslo.db, this
produces annoying warnings on MySQL 5.7 versions we use in the gate:
Warning: (3090, u"Changing sql mode 'NO_AUTO_CREATE_USER' is
deprecated. It will be removed in a future release.")
https://git.openstack.org/cgit/openstack/oslo.db/tree/oslo_db/options.py#n49
Unlike STRICT_ALL_TABLES, TRADITIONAL mode includes NO_AUTO_CREATE_USER,
and MySQL emits this warning on switching it on:
https://dev.mysql.com/worklog/task/?id=8326
So we have two options here:
1) make oslo.db default to STRICT_ALL_TABLES
2) make Devstack default to TRADITIONAL
The latter seems to be more appropriate as:
1) it's what modern MySQL versions default to
2) it's what people are actually using, if they do not override the
oslo.db default
3) it's more strict
Closes-Bug: #1652452
Change-Id: Ie6d823c9f8465ac9f2ce4825929d1a50438fab45
When cleanup devstack with linuxbridge, PUBLIC_BRIDGE should be
DOWN before trying to delete it.
Change-Id: I2d205cbe4d92a03ee5c376a23282d9880dd9a1df
Closes-Bug: #1662543
Tempest doesn't support the `api_version` config option for networking
anymore. I can't track which Tempest patch removed it, but it's been more
than 2 years.
Change-Id: I4012f470e8c317803203b6fa1e265600dbc49b3d
Commit e7361775c1 changed the code to
use Python 3 style print function, but when doing 'print()' in Python
2.7 it would print '()':
>>> print()
()
Import the __future__ print function so that a blank line will be
printed as expected. This will now work the same in Python 2 & 3.
Change-Id: I61742e107278f2327c18c9ab0de52d1914f16c97
When command failed and component failed to start, original exit code
was overwritten due to original command being executed in background.
This commit adds information about command's exit code to echoed message
about component's start up failure.
Change-Id: I8a3dd485b1b1f2d70d42c5610baac7c0c713f53a
Signed-off-by: Joanna Taryma <joanna.taryma@intel.com>
Now that Liberty is EOLed, the feature flag is not needed anymore.
Change-Id: Ib82cb21edbda383d17f8cf69fedc884f2357fead
Depends-On: I7073106988a79aad19c6b95bb050d2eaf00c36c0
The dedicated RPC worker is overkill in single or multinode
devstack deployments. Also metadata API workers was left
default, which meant they were as many as the CPU cores.
Related-bug: 1656386
Change-Id: Ibbf7787dfa48e13a51f961f3e0ee2b8f49964759
Right now we under pressure because of increasing memory consumption in
dsvm jobs. So it'll be good to see which process is eating the most ram
at a given time. It may not end up being useful, but it doesn't hurt to
at least display just in case.
Change-Id: I096bf4b425db51358240335e41f6238d1ec1bb40
While configuring the external network as the default
router gateway for IPV6 in lib/neutron_plugins/services/l3,
"router" keyword is missing in the command.
Corrected the command.
Change-Id: I055bea5137a841f709d4865ec9a43d6b53f8f4c9
Closes-Bug: 1660712
The map_cell0 command creates a cell mapping record in the
nova_api database, and the nova-manage db sync command
will migrate the db schema for the nova_cell0 database. This
patch takes advantage of that by moving the map_cell0 call
much earlier in the setup process so we get the nova_cell0
db schema migrated at the same time as the main nova db.
This also removes the || true condition around map_cell0
since it's idempotent now due to fix:
aa7b6ebbb254f00fcb548832941ca9dbd3996d9f
Change-Id: Ice4fbb1771270c618b2acbc933d4fbfb6805df81
Nothing uses this variable either in devstack or libraries,
so it's dead code (at least on master), and we can remove it.
Change-Id: I5975c476ae5b26402c209d6e5746e7a5a5a91507
Cells v1 apparently doesn't support the swap volume API which
was recently enabled for testing in change:
92575baa6b
Rather than revert that change, we should just handle the cells
v1 case and not enable that test in that environment.
Change-Id: I80f52e8299641098d90d3c374a80770fc45b8122
Closes-Bug: #1660511
tools/discover_hosts.sh is run by devstack-gate, and breaks all dsvm job
that doesn't use nova.
nova-manage is perhaps not installed if nova services are not enabled.
This change checks the presence of nova-.
Change-Id: Ic555d241f98d0fa027897c69a7115d1be88f6c96
This adds a simple script to run the
'nova-manage cell_v2 discover_hosts'
command which will be used by
devstack-gate to discover the compute
hosts after devstack is fully setup.
This allows us to manage the branches
where this can run from devstack rather
than require branch logic in devstack-gate.
Change-Id: Icc595d60de373471aa7ee8fb9f3a81fc12d80438
Depends-On: I4823737246a8e9cc4eaebf67ff6bdba8bf42ab29
If the NoVNC service is enabled, enable vnc_console in tempest.conf.
This will allow tempest tests that interact with VNC to be executed.
Change-Id: Idb38a3b11e2f61f23adf1ec23c04ddccd72e7539
Depends-On: I09aed8de28f1ba2637382e870134ced38808df29
The daemon mode of root helper for XenAPI has been implemented by
this change which has been merged to neutron:
https://review.openstack.org/#/c/390931/
It will help to import the performance. Let's enable this mode
by default in devstack.
Change-Id: I52246bef3e4434dfc49446535b122580bc475ac3
This provides a single setup_logging function which builds consistent
colorization if the config supports it, otherwise builds the identity
strings that we need to actually keep track of requests.
Change-Id: Iffe30326a5b974ad141aed6288f61e0d6fd18ca9
This makes setup_colorized_logging be a thing which takes a single
parameter and doesn't let projects do things differently. It also
changes the order of values from user / project to project / user to
represent the hierachy more clearly.
Change-Id: I8c0ba7da54be588e3e068734feb4f78ed7c5a14a
An initial install for devstack-tools, this will need to use all the
fun pip extra variables for installation, however the current
pip_install always prefers python2, and we only want to do python3
here.
Change-Id: I3dcdb35130f76fad81cb7b0d4001b7e96efbbd84
The test that is in tempest for this feature is
specific to LVM and will *not* work for other backends
regardless of them supporting the feature. It shouldn't
default to enabled for everyone, only for LVM.
If others want to opt-in they can, but its definitely
the minority that would.
Change-Id: I21347f2a5069059e6413208b254d5acd246faaea
In the incoming XenServer, it failed to install conntrack-tools
in Dom0 due to the bash script which is trying to find the correct
CentOS release version to be used in yum command. This patch is to
fix the problem
Change-Id: If7f169e118ccb7c29fc479c361417a916dc40b40
For some reason we were defaulting the name of the cell0 database
to nova_api_cell0 instead of nova_cell0. Devstack inherited that to
make things work, but we don't really want that. This patch makes us
use the proper name and create the cell0 mapping accordingly. As a
side effect, it also starts the process of unifying the cellsv1 and
cellsv2 paths by creating the cell0 mapping the same for both.
Change-Id: I4e7f6c5eaa068c98e5c4ef3feaee50d8e4f5d484
Findutils added in release 4.2.3 a new --delete action for deleting
matching files. This action performs better than -exec rm {} \;
because it doesn't have to spawn an external process. This change
uses a new action whenever is possible.
Change-Id: Iff16a86b18e924cfe78ac7c6107910940ce51e03
$ipv6_modes should always be passed when creating the
default IPv6 subnet, not just when fixed_range_v6 is
set. Without it the default was DHCPv6, which cirros
doesn't support out of the box. Was broken in
change-over from neutron to openstack cli.
Change-Id: Iadd39b1ce02fe0b3781bd3ae04adfd20d7e12d9f
Closes-bug: #1656098
Since cellsv2 setup is no longer optional, we can't even exclude
cellsv1 from this step. Since cellsv1 users can't use the simple
command, this does the individual steps as needed.
Depends-On: Icfbb17cce8ce8b03dc8b7b4ffb202db01e5218a6
Change-Id: I3c9101a34b2bb0804fc4deda62dbb8637e7b8f94
The only virt driver in nova that supports the swap volume API is
libvirt so enable testing that in Tempest only if using libvirt.
Depends on two changes:
1. The Tempest change that adds the new config option and test.
Depends-On: I2d4779de8d21aa84533f4f92d347e932db2de58e
2. A nova fix for correctly waiting for the block copy job in the guest
to complete.
Depends-On: I0c52917a5555a70c4973f37dea1aebf878dd73b4
Change-Id: Ibb6b309574d2c6a06fcecb0626ea21527fb7f412
Only a few Cinder backends support the 'manage snapshot' feature. So
we need a feature flag here. Luckily the LVM driver does support this
feature so default the feature flag to True in devstack(/Gate) but
introduce a variable to tweak the config.
Change-Id: Ifcb9f91059f08bdf2faf2a8d65229aba5742ee1c
Depends-On: I77be1cf85a946bf72e852f6378f0d7b43af8023a
In Nova, service token will be passed along with user token to communicate
with services when dealing with long running tasks like live migration.
This change addresses adding service user configuration for nova in
devstack.
Part of Nova blueprint use-service-tokens
Depends-On: I51eb0a8937fa39a2e5dafb1ad915e7113ea61f72
Co-Authored-By: Sarafraj Singh <sarafraj.singh@intel.com>
Change-Id: I2d7348c4a72af96c0ed2ef6c0ab75d16e9aec8fc
When installing a library from source and python 3 is enabled, first run
the installation process with python 2 enabled to ensure the library is
also installed under python 2 for any services not yet running under
3. The python 3 version is installed second so that command line tools
installed with libraries are installed under python 3 when python 3 is
enabled.
Change-Id: Ibb0f7a68d21081bf7652a0c1515080c0c54888ca
Signed-off-by: Doug Hellmann <doug@doughellmann.com>
Add variables ENABLED_PYTHON3_PACKAGES and DISABLED_PYTHON3_PACKAGES to
work like ENABLED_SERVICES and DISABLED_SERVICES and to manage which
packages are installed using Python 3. Move the list of whitelisted
packages in pip_install to the default for ENABLED_PYTHON3_PACKAGES,
except swift which is not enabled by default for now.
Add enable_python3_package and disable_python3_package functions to make
editing the variables from local.conf easier.
Add python3_enabled_for and python3_disabled_for functions to check the
settings against packages being installed by pip.
Update pip_install to check if python3 is disabled for a service, then
see if it is explicitly enabled, and only then fall back to looking at
the classifiers in the packaging metadata.
Update pip_install messages to give more detail about why the choice
between python 2 and 3 is being made for a given package.
Change-Id: I69857d4e11f4767928614a3b637c894bcd03491f
Signed-off-by: Doug Hellmann <doug@doughellmann.com>
Hypervisor XenServer will change to use os-xenapi in the future,
this will need DevStack changes, this patch is to remove install
Dom0 plugins part to our own DevStack plugins.
Change-Id: Ic327135b893a77672fd42af919f47f181e932773
The logic to set the default image for Ironic has been moved into the
Ironic tree. This patch is just removing it from DevStack.
Change-Id: Iaeb177f194adc83e40d86696e5553f9f72bbd1f9
Depends-On: Id828b41dc44113ce1cd094ce5fc245989699d4ff
The PCI-DSS feature has been introduced during the Newton
release and its settings are disabled by default. This
patch adds the possibility to enable some of them during
DevStack setup.
Change-Id: If6b5eb3e3cbc43eb241c94d18af80ad50be08772
Depends-On: Id97ca26f93b742cc3d8d49e98afc581f22360504
* iniuncomment followed by iniset for reseller_prefix just adds a
duplicate line in the config file that configparser does not like
so just remove the uncomment
* fall back to http:// url for glance->swift keystone authentication
* insecure flag to talk to swift
Depends-On: I51d56d16a5b175bd45dee09edc0b2748d72a5d06
Change-Id: I02ed01e20f8dce195c51273e8384130af53384ce
Currently if PHYSICAL_NETWORK and PUBLIC_PHYSICAL_NETWORK are
same then duplicate entry is created in ml2_conf.ini like below:
flat_networks = public,public,
With this patch, if PHYSICAL_NETWORK and PUBLIC_PHYSICAL_NETWORK
are same then add only PHYSICAL_NETWORK to flat_networks in
ml2_conf.ini
Change-Id: Iae4d1ee3882f6d96b4e4abd52ecc673a620563b5
Closes-Bug: #1654148
This commit switches how scripts we use to launch the installed version
in the path. Previously the scripts were manually executed in the source
repo, but this has issues if you're trying to run with py3 in a system
where python == py2. Setuptools already does the shebang magic for us
at install time, so we just need to use the installed version of the
script.
Change-Id: Iaa4d80ec607a2aa200400330e16cad3a4ca782ac
According to the feedback in the TC meeting [1], we renamed the Nova
virt driver from "docker" to "zun" [2] to avoid name collision
to nova-docker. This rename also help to clarify the difference
between these two drivers.
[1] http://eavesdrop.openstack.org/meetings/tc/2016/
tc.2016-11-29-20.01.log.html
[2] https://review.openstack.org/#/c/414651/
Change-Id: I747080953ae4d1d35ed334831100413b6e4466c4
The domain_name to be used needs to be $SERVICE_DOMAIN_NAME, as this is
changed in devstack from "Default" to "service".
Change-Id: I6351c1b2ca7ea4448e13eb87455bff4058df4fa7
Previous this was set the zake, but that was revert to missing
dependencies issue and because zake is a test fixture and not somthing
to deploy.
This change configures the Cinder dlm with this one is zookeeper.
And it installs tooz and the extra dependencies needed for the
zookeeper driver.
To do it, this commit have to introduce a new method for package
installation: 'pip_install_gr_extras package extra1,extra2'.
Change-Id: Idca310c08e345db59840eb31434c6cb1f849fa70
nova-network has been deprecated since Netwon and Nova change
I8388c29ad310cd8800084b4d5c026013158bfbed is switching the default
value of use_neutron to True, so we need devstack to explicitly
set use_neutron=False when running and configuring nova-network.
Part of blueprint use-neutron-by-default
Change-Id: I82721b5d10711401b9b0ebc2b0ed07cc8287bbf7
Really close to getting swift and keystone under uwsgi working, so
let's white list them. Won't affect any existing jobs, so we should
be good.
Change-Id: I51d56d16a5b175bd45dee09edc0b2748d72a5d06
Now that Liberty is EOLed, the feature flag is not needed anymore.
Change-Id: I5206535761773d4bcb02ebb8f25d1b0c1b59110c
Depends-On: If0b2168080a0b0ecdc6682ef69856a0879f4f6d3
CLI tests have been removed from Tempest in
I4f8638f1c048bbdb598dd181f4af272ef9923806
Dashboard tests have been removed from Tempest in
I2a69ebed2947a5ab5e5ca79557130bd093e168dd
Change-Id: I6df74a07e209b07fd3feae762c9cdab16e09414f
The snapshot_backup feature flag was introduced in
Ib695e60c2ed7edf30c8baef9e00f0307b1156551 to enable Tempest tests
introduced in I1964ce6e1298041f8238d76fa4b7029d2d23bbfb
But I1964ce6e1298041f8238d76fa4b7029d2d23bbfb was never merged so that
feature flag was never really useful.
Change-Id: I4e0bc786d2320907cb101fc788ad51444628537d
Use trueorfalse to normalize the values for USE_PYTHON3
Install 3.5 instead of 3.4 When USE_PYTHON3 is specified.
Also, since not many packages are classified correctly, fallback
to looking for just "Programming Language :: Python :: 3" and
log a message for the package to highlight the problem.
Also special case some services that are *almost* ready
Depends-On: Id48e1b328230fcdf97ed1cb4b97f4c3f9cf6eb8a
Depends-On: Ib7d9aa0e0b74a936002e0eea0b3af05102b06a62
Change-Id: I243ea4b76f0d5ef57a03b5b0798a05468ee6de9b
In order to test whether live migration is backward compatible
we need to live migrate VM back and forth between two versions
of nova. This configuration option will allow to reuse existing
tests just by adding if condition in a method that invokes live
migration and validates outcome:
* If set to False, it will use existing behaviour
* If set to True, it will live migrate VM, validate whether it
succeded, then live migrate the same VM once again and again
validate the result
Depends-On: Icaeca404ec3e4b8f3cd489789fdac6117740ec43
Change-Id: I8da2b3bd0c08d9a3111d3531c346d06bd52cae7b
The placement API configuration was binding a specific port *and* was supporting
to be called by the default HTTPd ports using a Location directive.
Given that the corresponding service catalog entry for the placement service type
doesn't mention the specific application port but is rather using the default
port 80, we can remove that specific port and just use the default config.
Note that we still need to use a VirtualHost directive for the specific placement
config because ErrorLog is only scoped for either server or virtualhost but can't
be set for a Location (or a Directory) context.
Change-Id: I9a26dcff4b879cf9e82e43a3d1aca2e4fe6aa3e6
This moves setting of RABBIT_HOST from stack.sh to lib/rpc_backend
so it may be used in grenade runs, which don't have the defaulted
value from stack.sh. The RABBIT_HOST is needed in order to call
get_transport_url in lib/rpc_backend.
Change-Id: I504f7fac7bb9a8c158e20046dbd1dd2d507db02b
Closes-Bug: #1649586
Depends-On: I3d4d7b309e50f4e2970cda55aada02d68c4fa705
Services that run inside Apache use tail -f on the corresponding log
file to display output in the screen session. However, they also use sed
to replace some control characters, and this means that the output is
buffered. This results in debugging experiences that range from
"impossible" (the log you want isn't shown) to "Kafkaesque nightmare"
(the log you want isn't shown, except that sometimes it is, and
sometimes it isn't even though you double-checked and you're completely
sure that you must have output a log, but when you check back later you
realise it actually is and you wonder if history is actually not mutable
after all and begin to question what is real and what is not).
This adds the --unbuffered option to ensure streaming output.
Change-Id: I665ff5f047156401d8152f478d834ac40ff31658
Now that the placement service is mandatory for running Nova in Ocata,
we want to enable it by default when running devstack by default.
In the past, we added a placement-client service with
I04a655fbc58913b3d607400a7f677be299499142
Devstack-gate will also be able to run a multinode devstack with the
help of Ibd760c642e3c1ffff2dd61be48e30530b0d24720
Change-Id: I273c3c8299ee329bed425f3e7cd4b583ed1187a4
Neutron doesn't use any arptables based firewall rules. This should
somewhat optimize kernel packet processing performance.
I think the setting came from:
http://wiki.libvirt.org/page/Net.bridge.bridge-nf-call_and_sysctl.conf
but does not apply to the way we use iptables.
Change-Id: I41796c76172f5243e4f9c4902363abb1f19d0d12
Closes-Bug: #1651765
If 'RECREATE_KEYSTONE_DB=False' database won't be recreated.
It would be useful for multinode Grenade tests for Keystone. This parameter
will help us to deploy multiple services on different machines talking to
the same DB.
Devstack recreates Keystone DB each time during Keystone service installation.
If our controller node is already deployed, Keystone DB already contains
important information about OpenStack services and their endpoints. When
the second Keystone node is being deployed, we don't want to delete
records about controllers' services endpoints.
Partially-Implements: bp rolling-upgrade-testing
Change-Id: Ia8d07b4295ca165be01e44466c95d5275f596e83
When doing multinode devstack we need a way to specify that we've
enabled for the placement service. We use a pseudo service of
placement-client for this.
Change-Id: I04a655fbc58913b3d607400a7f677be299499142
The current code assumes that there exists a public openstack network
and uses that assumption to set the public_network_id variable in
tempest lib. If NEUTRON_CREATE_INITIAL_NETWORKS is set to false this
step will fail as there is no public network to be found. This change
adds a check for NEUTRON_CREATE_INITIAL_NETWORKS before attempting to
set this variable.
Change-Id: I62e74d350d6533fa842d64c15b01b1a3d42c71c2
Closes-Bug: #1645900
Different versions of Ubuntu ship with different versions of Java.
Trusty had 7, Xenial has 8, and so on. This causes problems when we
hardcode a versioned package name into our dep lists as that version may
not exist everywhere. Thankfully Ubuntu provides a default-jre-headless
package that we can use instead that maps properly onto whatever java
version is correct.
Change-Id: I4e5da215c8f7aa426494686d5043995ce5d3c3af
* Set ml2_type_flat setting so that the public flat network is created
correctly
* Set securitygroup driver correctly
It should be set as:
[securitygroup]
firewall_driver = iptables
Change-Id: I7369b45fbc5a47ce958693c67a1902a8cb24f367
When creating a service user we allow the user to be created with a
different role. Currently in auth_token middleware we want to check that
the service token is specified with the service role so we should always
add the service role and optionally add additional roles.
Change-Id: Ie954a679674b4795079b539ebc8d4d2dcbd7dacc
When using local.conf in multinode envs not everything is going to be
defined in all places. Eventually we probably want to make it so we
have a host role for these sections or something. But for now warn
instead of die when we can't find a config var.
Change-Id: I6959099373f035fbfe9e540a44e4c52b8e7c95c0
Closes-Bug: #2000824
Since I21ae13a6c029e8ac89484faa212434911160fd51 nova-manage db sync
may try to make a request to api db in order to get cell mapping and
will fail, as the db is not created yet. While this is non fatal, we
could avoid the error anyway.
Change-Id: I19483e9420071d484f029779bcc8c6d623c210ce
Related-Bug: #1631033
the python-xml is a subpackage from the standard cpython package that
that contains elementtree and other bits that are needed almost
everywhere in OpenStack but isn't installed on a absolutely minimal
openSUSE Leap installation. This package doesn't exist on pip but
is a SUSE only invention, so just treat it similar to a bindep.
Change-Id: I82887c2e6895740d1b16d1269574519450ca783e
Starting with SLE12 SP2 and with openSUSE Leap the distro-shipped
openvswitch is the normal systemd openvswitch.service service file
and no longer the older openvswitch-switch Sysv5 init script. Add
a special case for that.
Change-Id: I5152f2585c3d4d18853988d6290039d6b1713b99
Now all configuration are present in Ironic tempest plugin
and those are going to be removed from tempest in
Id518a6d87d0949737cd1c50cb6a83149b85e5f85
Patch- I73c649625d106fc7f068e12e21eaacba8f43cbbb set
those in ironic devstack plugin.
We can remove all baremetal config setting from devstack.
Along with moved one this patch deletes other unused baremetal
config setting.
Change-Id: If826321ebc0c20ea372d206d49383f3826c9b547
Depends-On: Id518a6d87d0949737cd1c50cb6a83149b85e5f85
Depends-On: I73c649625d106fc7f068e12e21eaacba8f43cbbb
On Ubuntu nodes, devstack tries to predefine the initial mysql root
password by doing some debconf-set-selections, but these will not take
effect if the corresponding package has been installed earlier. So
just try to set it every time, like we do on other distros.
Change-Id: I2c167051fc5e53dd0ccf82a60ab085cd9cdea28d
create_cell requires n-api and at least one n-cpu up and running. If
we have a configuration where it is not guarunteed that there is an
n-cpu at the end of a devstack run we have to skip this step and make
the user run it manually later.
Change-Id: I2287ab29f3c1a7252271dcce81673ef365615296
Role "root" it is hardcode.
In general case role name comes from local.conf: string "DATABASE_USER="
Change-Id: Iedfca48e04d23c313851f48d68ac40ba29340805
The linuxbridge agent for Neutron expects that the public bridge will
already be created by the time it starts. On devstack, this only occurs
as part of the l3 agent configuration. If a compute node doesn't have an
l3 agent and is using a linuxbridge agent, then br-ex won't be created
and the process will not be able to start (causing stack.sh to fail).
This causes the gate-grenade-dsvm-neutron-linuxbridge-multinode-nv gate
to fail. To avoid the issue, skip the bridge mappings setup unless L3 is
configured. This is done in a backward compatible fashion: if localrc
uses the old q-l3 tags, the is_service_enabled neutron-l3 would not be
able to succeed.
Closes-Bug: #1643562
Change-Id: I292ff0dc080fb84b5f879ba2f00f03eff295b55b
libguestfs does not work on ubuntu because the kernel is not
world readable. This breaks file injection with libvirt.
See https://bugs.launchpad.net/ubuntu/+source/linux/+bug/759725
for more details.
The workaround proposed by Ubuntu is to relax the kernel ACL
if needed, so we need to do that in case file injection is
enabled on an Ubuntu host running libvirt.
Partial-bug: #1646002
Change-Id: I405793b9e145308e51a08710d8e5df720aec6fde
We need a local.conf processing phase after every run phase which lets
us override config options after that point. We didn't explicitly
support this for test-config before, which broke some CI systems when
we moved tempest to use this later phase.
Closes-Bug: #1646391
Change-Id: I7d693afa19acf3e8231e84e45b7a868628ebdbc0
This single global variable is no longer useful as we have multiple
repositories and devstack plugins nowadays.
Also, add a utility function, neutron_server_config_add, for devstack
plugins to add an extra config file.
Related-Bug: #1599936
Change-Id: I90112823ef96ae2fba97d7b09b00bec8cb816d8d
So that it can be used by functions like _determine_config_server,
which is used like RESULT=$(_determine_config_server).
Change-Id: Ia4e641c5529b95ada30ae662221f370bc7fa88a7
To avoid using legacy functions accidentially.
Depends-On: Ida1f83b6b3ef9b76be13c063c7e35a8703214078
Change-Id: I3ff136fc8330c92007cdfe91b77d7f9865eabd8d
XenServer already support OVS native mode and I have a patch for
configuring it https://review.openstack.org/#/c/372952/ which
is fine. But we have another patch which revert the usage of
ml2_confi.ini and ml2_conf.ini.domU
https://review.openstack.org/#/c/396573/. Both patches work well
separately. But the two should have some dependent relationship.
Once one merged, the other should change accordingly. Sorry that
we missed the dependency.
This patch is to fix the ovs config based on reverted ml2_conf.ini
and ml2_conf.ini.domU to make sure we configure the correct IP for
ovs agent
Change-Id: Ib53e37e210cc849f161dd6630f81e5b2331a91d5
Similar to 30ab23cd9b, fix the
plugin name to avoid warnings like:
WARNING stevedore.named [-] Could not load neutron.plugins.ml2.plugin.Ml2Plugin
Change-Id: Ibb45f1305816b255ba2419ba662d9e29eff68f58
Neutron has changed to use ovs native interface by default, but when
the hypervisor is XenServer, we cannot use ovs native interface without
extra configurations in neutron-openvswitch-agent(q-agt) in compute
node.
This patch is to add the needed configurations automatically during
deployment, so user needn't to do it manually and restart q-agt.
Change-Id: Ibc69d3cdb4d75833f2ac16840c62bcacf460dd4f
We are seeing connection errors to the proxy occasionally. These errors
do not result in a logged http request or error to the backends,
resulting in a theory that the proxy itself may just not be able to
handle the number of connections. More than double the total number of
connections that will be accepted by the proxy in an attempt to fix
this.
Change-Id: Iefa6c43451dd1f95927528d2ce0003c84248847f
Related-bug: 1630664
There may be cases when the configuration of the OVS is different
from the default one. This enables one to make use of the neutron
configuration file to contain all of the OVS settings.
Change-Id: I728cf8cdc653667c076b07b39c13c1278281c01b
Closes-bug: #1645691
The order of the steps were a bit confusing for the first timers
in the devstack document. So, changed the order of installation
steps to make it clear.
Change-Id: Ifaa051887dab95719b9ca5d1b2fbe2f5f549d269
Closes-Bug: #1627939
The current coding fails to process local.conf like
the following. Note: This example is taken from a
real use case. [1]
[[post-config|$NEUTRON_CONF]]
[qos]
notification_drivers = midonet
[[post-config|$NEUTRON_CONF]]
[quotas]
# x10 of default quotas (at the time of writing)
quota_network=100
quota_subnet=100
quota_port=500
quota_router=100
quota_floatingip=500
quota_security_group=100
quota_security_group_rule=1000
[1] https://review.openstack.org/#/c/400627/
Closes-Bug: #1583214
Change-Id: Ie571b5fa5a33d9ed09f30ba7c7724b958ce17616
Horizon is removing run_tests in favour of tox during Ocata, as part of
https://blueprints.launchpad.net/horizon/+spec/enhance-tox. To complete
this move, we need to remove any reliance on run_tests.
Change-Id: Ia8ad073aee68d1660d3bb5a68ec07516d8ce0665
In the 'Private Network Addressing' section of the doc,
there are references to FIXED_RANGE when referring to V6
networks. These have been changed to FIXED_RANGE_V6.
Also fixed a few typos and grammatical errors when
giving the doc a quick read-through looking for more
references to FIXED_RANGE.
Change-Id: Iaa530c476ce2b36a3f616945ddd2e24fa599a16c
If devstack is deployed in the VM with defined
public IP address (like 192.168.10.6) it is not possible to
access the Horizon from the browser.
This is because DEBUG=True means that ALLOWED_HOSTS, if not set,
is equal to ['localhost', '127.0.0.1', '[::1]'] according
to Django's documentation.
Change-Id: I0ab2b57e459dbfa3b01b3e9388bbcefac076a142
Co-Authored-By: David Lyle <dklyle0@gmail.com>
Closes-Bug: #1643050
Using devstack on a RHEL based system results in
"Listen 0.0.0.0:80" being added to the
/etc/httpd/conf/httpd.conf.
This configures Apache to only listen to port 80 on an IPv4 interface.
This makes it not possible to access Horizon via IPv6 without
re-configuring and restarting httpd.
Removing this sed leaves the default "Listen 80" from the rpm package,
which binds to all interfaces and will allow connection to Horizon
via IPv6.
Change-Id: I9fe8cbebff0ca6a30ceeaae0f7e035c9bb828d44
No need to explain how to enable Neutron now that it's enabled by
default. Keep but reformat the 'how to enable swift' part though.
Change-Id: I3f9b7796fad10abf1039e4c68eb2cd5ef6cdbc99
Since 4b49e409f8 devstack
started to use reload instead of restart.
Using reload in devstack for a fresh install,
does not makes too much sense unless multiple service
plugin touches the same service configs.
Systemd rejects to reload something,
which was not loaded before.
$ sudo /bin/systemctl reload httpd
httpd.service is not active, cannot reload.
We will switch to `reload-or-restart` action instead of `reload`,
it is more likely the action what the previous patch wanted.
Change-Id: I70d597fbe4a8923d937ba8432e29edefb27d1058
The linuxbridge agent for Neutron expects that the public bridge will
already be created by the time it starts. On devstack, this only occurs
as part of the l3 agent configuration. If a compute node doesn't have an
l3 agent and is using a linuxbridge agent, then br-ex won't be created
and the process will not be able to start (causing stack.sh to fail).
This causes the gate-grenade-dsvm-neutron-linuxbridge-multinode-nv gate
to fail.
Closes-Bug: #1643562
Change-Id: I6f441c6febb5070ad885569d9c798634d0272b6c
Change dddb2c7b5f recently
changed devstack to enable the Cinder image cache by default
and changed to use thinly provisioned LVM volumes by default.
Since then we've had a spike in thin LVM snapshot test failures
in the gate, which is by far our top gate bug at 219 hits in the
last 10 days.
So unless there is a fix on the Cinder side, this changes the
default lvm_type back to 'default' for thick provisioning.
Change-Id: I1c53bbe40177fe104ed0a222124bbc45c553b817
Related-Bug: #1642111
With the key manager refactoring in nova and cinder, the key manager
class will need to be explicitly set.
Nova key manager refactoring: Ib563b0ea4b8b4bc1833bf52bf49a68546c384996
Cinder key manager refactoring: Ief8885bb4ca8d62b03cf1a52c25dd0e62c835bfe
Change-Id: I733279864ee1a4aaffc9c8eed81b5e12f8d8821b
Implements: blueprint use-castellan-key-manager
Without this parameter, when we set GIT_DEPTH,
it may happen that we clone only master and
then cannot checkout branch
Change-Id: I39376914f8bfc286a308c99db6bc92cddab195b5
The intent was to make any ipv6 safe addr range bigger than a /64 a /64
when setting the fixed range. Unfortunately the awk only emited the mask
and not the addr. Fix this by sprinkling the address back in.
Fixes-Bug: 1643055
Change-Id: I526d4c748fd404ecb3c77afcbb056aa95090c409
If devstack is deployed in the VM with defined
public IP address (like 192.168.10.6) it is not possible to
access the Horizon from the browser.
This is because DEBUG=True means that ALLOWED_HOSTS, if not set,
is equal to ['localhost', '127.0.0.1', '[::1]'] according
to Django's documentation.
Change-Id: I74ae99569dafa10eee7066713a05fb49183e3fca
Currently devstack assumes that the network used for ssh
validation is the private network. This patch adds a hook that
sets the network used for ssh validation based on whether or not
provider networking is being used. It also moves the function
'is_provider_network' into functions-common as it will now be
used by both tempest and neutron.
Change-Id: I265c9e26c9bfb18b7e201f27d8912b8bec235872
We should be using $VOLUME_GROUP_NAME instead since Icehouse.
$VOLUME_GROUP_NAME has been introduced in
I93b8ef32832269d730c76a6dc24ddb4f20c6d9df and $VOLUME_GROUP is nowadays
only use as a fallback to $VOLUME_GROUP_NAME.
As a code comment in lib/lvm says it we kept the $VOLUME_GROUP around as
"for compatibility with icehouse-generation Grenade". Icehouse is long
gone so now seems a good time to remove any usage of $VOLUME_GROUP.
Change-Id: Id3051b5a196c45266c39fde4f08401aaacf0f6bd
When using zypper remove, include the -y option to avoid stack.sh from
hanging waiting for user confirmation. Due to output buffering, the
script could hang before giving the user the prompt to enter Y to
continue, making it unclear why the script was hanging.
Change-Id: I5ea761e5ae0829439953c385f8e7d0546acba886
Closes-Bug: 1642736
Swift port base was changed in Ifd95b99004aead5ddc8ae1a8dd3ccd9c4f2abe91
but we forgot to update the rsyncd.conf. This patch update the rsyncd.conf
file.
Change-Id: Id457c047c672a810c4c0c7721b6beeb01b719879
When using the enable_plugin command and grenade jobs it can be
easy to enable the same plugin twice, as the grenade job has a
registration section and the configuration in project-config can also
enable it due to code-reuse in project-config.
If a plugin is enabled twice it will likely fail, though it won't be
obvious that it was due to the plugin being enabled multiple times.
This change makes it so if it sees the same plugin name is enabled
more than once it will die and an error message outputted.
Change-Id: I9f1d7e58b861b04473b6a57c9ad404203fb7277a
The switch to using subnetpools caused quite a bit of confusion
because it didn't respect the value of FIXED_RANGE. This caused
conflicts in the gate with it's default IPv4 value of 10.0.0.0/8.
This patch does a few things to address the issue:
* It introduces the IPV4_ADDRS_SAFE_TO_USE and IPV6_ADDRS_SAFE_TO_USE
values and adjusts all of the FIXED_RANGE and SUBNETPOOL_PREFIX values
to dervive from them by default.
* This addresses the concern that was raised about implying that
SUBNETPOOL_PREFIX and FIXED_RANGE are equivalent when setting
SUBNETPOOL_PREFIX=FIXED_RANGE by default. Now we have a new value
for the operator specify a chunk of addresses that are safe to
use for private networks without implementation implications.
* Backwards compatibility is maintained by alloing users to override
override all of these values.
* The default for IPV4_ADDRS_SAFE_TO_USE uses /22 instead of /24
* Because we want to be able to use subnetpools for auto allocated
topologies and we want to be able to have a large chunk of
instances on each network, we needed a little more breathing room
in the default v4 network size.
* SUBNET_POOL_SIZE_V4 default is changed from 24 to 26
* In conjuction with this change and the one above, the default
subnetpool will support up to 16 64-address allocations.
* This should be enough to cover any regular gate scenarios.
* If someone wants a bigger/smaller subnet, they can ask for that
in the API request, change this value themselves, or use a different
network entirely.
* FIXED_RANGE_V6 defaults to a max prefix of /64 from IPV6_ADDRS_SAFE_TO_USE
* This avoids the private subnet in the non-subnetpool case from being
larger than /64 to avoid issues identified in rfc 7421.
* Users can still explicitly set this value to whatever they want.
This 'max' behavior is only for the default.
* This allows IPV6_ADDRS_SAFE_TO_USE to default to a /56, which leaves
tons of room for v6 subnetpools.
Closes-Bug: #1629133
Change-Id: I7b32804d47bec743c0b13e434e6a7958728896ea
This removes the logic to add a route pointing to the IPv4
tenant private network range since the router is performing
SNAT. If reaching the IPs via the route worked at all, it was
by accident since this behavior is certainly not guaranteed
by Neutron.
Change-Id: If45e3fc15c050cfbac11b57c1eaf137dd7ed816f
XenAPI requires two instances of L2Agent: the standard one manages OVS
bridges in DomU and the service name is called as q-agt in Devstack;
the other new L2Agent manages OVS bridges in Dom0 and the service name
is called as q-domuA. In order to support the new agent q-domuA, it
requires some XenAPI-specific configurations. But unfortunately those
XenAPI-specific configurations were configured in the standard agent
file, meaning other changes made to the standard agent file would not
have the correct effect. So it has caused issues, for example, floating
IP addresses are not reachable.
This fix is to move the XenAPI-specific configurations from the stardard
agent configuration file to the XenAPI-specific agent configuration file
so that it won't impact the standard agent's behavior.
Change-Id: I45944e84a1f81d016aa00da6d782801ee8457ea4
Currently the x509 certificate setup is done after all the
openstack services have been deployed. This is OK because
none of the services require that the x509 certs exist
when they are being deployed. With the integration of TLS
into the nova novnc proxy (and later spice & serial proxy)
service, x509 certs will need to exist before Nova is
deployed.
The CA setup must thus be moved earlier in the devstack
deployment flow, prior to the setup of any services. One
part of the CA setup, however, fixes up the global cert
bundle locations and this can only be done after the
python requests module is install, thus must remain in
its current location.
Change-Id: Idcd264fb73bb88dc2f4280c53c013dfe4364afff
This removes all of the heat code from the devstack tree, in favor of the
devstack plugin in Heat's tree.
Depends-On: I4bed1e5cef5afa7b049b07640086a86a3f881e13
Depends-On: Ic392bcc24bc374ee8511a94f1d8f6ac23131c7e3
Change-Id: I5b60422bf1f5fa78aa8f3383f7a222e0356d9e42
The deprecated AMI image file opts will be removed soon.
See https://review.openstack.org/#/c/338377.
So we can't use the fallback mechanism anymore. This patch is to
specify the correct image parameters for XenServer.
Change-Id: Ic287a3ed1725c42ea29022158bc9720c9a96533f
Previously the usage of neutron debug ports was removed by
5e01c47e4d but there was still call to
teardown_neutron_debug. Recently a change to devstack-gate
1d6cc0771a3399300117f488e9d71e7ea46a4d82 caused that call to be
triggered and breaking the gate-devstack-dsvm-updown job.
This patch deletes the call and comments regarding setup_neutron_debug
and teardown_neutron_debug.
Change-Id: Ifdacb0cec1307db469bd66f551474539184cf2cd
Besides updating to OSC CLI, this patch also fixes an argument name typo
present before in 'nova keypair-add' (--pub_key should be --pub-key).
Specifying $OS_PROJECT_NAME in case user is associated to multiple
projects containing security groups with same name (e.g. 'default').
Change-Id: I776f6edfc4c6c798a39d3260827a18c695f05c87
Nova is going to land a database migration in Ocata
under change I72fb724dc13e1a5f4e97c58915b538ba761c582d
which enforces that at least the simple cells v2 setup
is performed, which creates the cell mappings, cell0 and
host mappings. Before we can land that change in Nova
we have to make cells v2 setup a default in the integrated
gate jobs.
Depends-On: Ie44e615384df464516aa30b9044b5e54b7d995bb
Change-Id: If1af9c478e8ea2420f2523a9bb8b70fafddc86b7
The neutron client is going to be deprecated during the
Ocata timeframe, so it is time to start switching to the
openstack client to invoke networking commands.
use of neutron client in neutron-legacy has been left as is.
The command for setting the router gateway is left as follow up.
Change-Id: I0a63e03d7d4a08ad6c27f2729fc298322baab397
When using neutron network under xenserver, we must enable linux bridge
in Dom0 as neutron will use linux bridge qbr in compute node for
security group. But by default XenServer use openvswitch and disabled
linux bridge. This patch is to remove this restriction.
Change-Id: I0e8124ff2323810fdc46c717a750ce7e8f4aa0c6
The initial start of the neutron OVS agent always prints
a warning:
WARNING stevedore.named [] Could not load
neutron.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver
There's an alias for that in setup.cfg called
iptables_hybrid that would avoid it.
Change-Id: I3f5bf782f4f27dc123e462e494741a8a941641ec
This patch sets Keystone v3 as default in services
configuration files and in the openrc and stackrc scripts.
Change-Id: I24546f02067ea23d088d383b85e3a78d7b43f166
Partially-Implements: bp keystonev3
When using XenServer, it will create two neutron-openvswitch-agent
q-agt and q-domua even it's single box environment, but it didn't
stop the q-domua, this patch is to stop q-domua in unstack.sh
Change-Id: I511ed534bfb7d5fe6136f6a0b33f1d749d30862c
Closes-Bug: #1631721
This commit removes some config values for tempest that no
longer exist in tempest/config.py therefore are no longer needed
in tempest.conf.
Change-Id: I5778973012e57e8d9df9bf864590f8ed7fe05561
Sets the port_security feature flag in tempest.conf
if the port_security extension is enabled, which it's not
by default in neutron but is set by default in devstack.
This adds global variable for setting the port_security
extension in ml2.conf and in tempest.conf so we only have
to set this in one place.
Depends-On: I1efd5c838aa0d73cc6e8864e3041eea25850198d
Change-Id: I6334b200e42edd785f74cfb41520627393039619
Related-Bug: #1624082
this is the first patch in a series to actually make fernet the default
token provider in keystone. the patches for grenade, release notes, and
actually switching the value in keystone all depend on this patch first.
reasons for switching over:
- fernet tokens are the recommended token provider
- the install guide for newton recommends deployers use fernet tokens [0]
- we previously attempted this switch but ran into timing issues [1],
the timing issues have been resolved [2]
[0] http://docs.openstack.org/newton/install-guide-ubuntu/keystone-install.html
[1] 153db26970
[2] https://review.openstack.org/#/q/topic:make-fernet-default
Change-Id: I3b819ae8d2924f3bece03902e05d1a8c5e5923f1
The devstack ldap configuration for keystone is still using some
old options that are no longer valid. The write support is
being removed this release. And in previous releases, the ldap
assignment driver support was removed and was not removed here.
Change-Id: I538626b681eaee6a7ac10dfbc29605b73fbe13bf
To avoid it being created multiple times for multinode setup.
Note: This reverts "Enable neutron to work in a multi node setup"
(commit 88f8558d87) partly and fixes
the issue differently.
The configuration in question uses the new lib/neutron. (not neutron-legacy)
In that case, calling create_neutron_initial_network from stack.sh directly
is a wrong way, as create_neutron_initial_network is sourced by
neutron-legacy. The new neutron code should not rely on the legacy one.
Closes-Bug: #1613069
Change-Id: I868afeb065d80d8ccd57630b90658e330ab94251
Q_ variables belong to neutron-legacy.
These are True by default in neutron.
Remove them in favor of post-config meta section.
Change-Id: If691a79b09003f85a07c9f33e0379a2b21e48141
Until the policy changes land for Nova, Glance, etc, this
value is not used. Additionally, by having it set, it actually
makes it hard/impossible for the required changes to land in
the other services. Disable/comment out the changes in the
Keystone specific lib file for now, and we will re-enable once
the Services can make use of them.
Change-Id: Ia1de9083c21107dac2f0abb56bda166bdb37a69d
The ceph cinder backend script was setting the wrong
config option in cinder.conf for the secret uuid. This
was being masked by a bug in nova which is failing on
this bug when trying to fix the nova bug...right. It
makes sense.
See:
http://docs.ceph.com/docs/master/rbd/rbd-openstack/#configuring-cinder
Change-Id: I4655cae3212d589177d2570403b563a83aad529a
Closes-Bug: #1635488
It's not used, and a recent change to trim down projects lists in
devstack-gate broke devstack in the gate that enabled heat.
Change-Id: I405423bdc9ba8dd9b30fce6fdceacccf662d5da3
Ubuntu wily support is EOL so lets make room for yakkety.
Change-Id: Ib13d43f6d89bdf7c684cd34655a077a13e237be3
Signed-off-by: Chuck Short <chuck.short@canonical.com>
This reverts commit 6930ba312f.
By reverting this patch we are no longer using the bandaid fix mentioned
in the code. The latest openstackclient release (3.3.0) fixes the bug.
Related-Bug: 1619274
Change-Id: I20e3c5a92b97bf46c8d2318cd37044f0f36e1745
We should not require any special role for heat
since very long time.
We should use the same roles as with the primary user.
Change-Id: Id9150f94c30505ed0da33b8fbc2a5a7bd4fcf5d0
When tls is enabled, we aren't bringing the logs to the forefront,
which makes it hard to debug when things go wrong. This does that.
Change-Id: I7c6c7e324e16da6b9bfa44f4bad17401ca4ed7e3
The prior art on other options in the same document seemed to be
calling out the default in a pre-formatted block after describing the
possible values.
I believe the default value for the option was first changed [1], then
the docs were fixed [2], then the information was unintentionally
dropped from the docs [3].
1. Related-Change: If0e0b818355e4cb1338f7fa72af5e81e24361574
2. Related-Change: Ib6603b4f6ea0b4079f9a4ea46e723ecbb2ea371d
3. Related-Change: Iddd27cb54f1d9f062b9c47ff9ad6a2bef3650d6b
Change-Id: I662403db3b08a351a680587440ad1f15a6f8ee5d
doing a clean.sh / stack.sh cycle with USE_SSL=True was failing
because we were no longer cleaning up the keystone site fully, so some
of the early mod_ssl queries hit an invalid apache configuration.
Change-Id: Ic6f3f601e532ec50c0234d928c25b378d9e95e32
This fix replaces Q_USE_PROVIDERNET_FOR_PUBLIC with
Q_USE_PROVIDER_NETWORKING in the error messages introduced by
[1].
The error is thrown when provider networking with IPv6 has been
requested via local.conf, but no provider IPv6 range or provider
IPv6 gateway is provided. But if a provider network should be used
over the private network is determined along the variable
Q_USE_PROVIDER_NETWORKING and not Q_USE_PROVIDERNET_FOR_PUBLIC.
The variable Q_USE_PROVIDERNET_FOR_PUBLIC determines if a provider
network should be used as public network. This happens a few lines
later in the code and is not related to those error messages.
[1] https://review.openstack.org/#/c/326638/1/lib/neutron_plugins/
services/l3
Change-Id: I50aa1e9d2027eef598c95404851e51c31a397fbb
This creates log files per proxy vhost and sets the log level to info to
help debug potential issues with tls proxying.
Change-Id: I02a62224662b021b35c293909ba045b4b74e1df8
If DLM is enabled, cinder should be configured to use the correct
backend url for the dlm.
At the moment only zookeeper is supported, as it is the only backend
currently supported in devstack.
Change-Id: I7afc8dc95bc5b3f11b888e10607615c1212c45f4
As long as nova already supports an Identity v3 auth flow when talking
to ironic (Id837d26bb21c158de0504627e488c0692aef1e24), make it use
v3 by default.
This way we don't fail in a keystone v3-only situation, for
example.
Change-Id: I028dfb52108d0630f47a53f8b420b70d4979eb55
Some of the clouds used for CI use the 10.2xx.0.0/16 range
for VMs, and collide with the wider 10.0.0.0/8.
This setting allows for creation of 256 subnets out of the pool.
Change-Id: I48c86f94098f1501f0e7f90a265dda7e81440eb0
Closes-Bug: 1629133
Added an option to make subnetpools to be optional
as it ignores the public network specified in
FIXED_RANGE.
DocImpact
Change-Id: Ic89ceca76afda67da5545111972c3348011f294f
Closes-Bug: #1628267
With the plan [1] to stop enabling it by Neutron iptables firewall
driver itself, deployment tools should catch up and enable the firewall
themselves.
This is needed for distributions that decided to disable the kernel
firewall by default (upstream kernel has it enabled). This is also
needed for distributions that ship newer kernels but don't load the
br_netfilter module before starting nova-network or Neutron iptables
firewall driver. In the latter case, firewall may not work, depending on
the order of operations executed by the driver.
To isolate devstack setups from the difference in distribution
kernel configuration and version, the following steps are done:
- we load bridge kernel module, and br_netfilter if present, to get
access to sysctl knobs controlling the firewall;
- once knobs are available, we unconditionally set them to 1, to make
sure the firewall is in effect.
More details at:
http://wiki.libvirt.org/page/Net.bridge.bridge-nf-call_and_sysctl.conf
[1] I9137ea017624ac92a05f73863b77f9ee4681bbe7
Change-Id: Id6bfd9595f0772a63d1096ef83ebbb6cd630fafd
Related-Bug: #1622914
Nova ships with an empty policy.json file which it does not need.
oslo.policy previously required the empty file to be there but as of
version 1.14.0 it is possible to run with no policy file at all. Since
there are no policies defined in the sample file let's no install it.
Change-Id: I85a251376dfe38caa4b100861bf764014a98bc37
Depends-On: I09fa842ffbe75bed269cef6edc9c82d18bfe9297
When dots are used with sysctl, they are reinterpreted as slashes.
Route devices can have dots in their names, so when they are used in a
sysctl command that also uses dots, its dot will be replaced with a
slash, causing an error.
Change-Id: Ie32126a3aa8d646568d7d37ec4874419b9658935
Closes-Bug: #1627770
The motivation is to make it more friendly with lib/neutron.
ie. independent from lib/neutron-legacy
Change-Id: I19821b009cbf1bc715a6c7b2854e4c77d2041ec4
urllib3 1.18 was released today and contains new more correct hostname
matching that takes into account the ipAddress portion of a certificate
and disallows matching an IP Address against a DNS hostname.
Change-Id: I37d247b68911dc85f55adec6a7952ed321c1b1d8
The /identity_admin endpoint is the port 80/443 equivalent of the
service that typically runs on port 35357. In v2 some operations must be
performed on the admin endpoint whereas on v3 the services on 5000 and
35357 are exactly the same. This would be why the service was mounted at
/identity_v2_admin however that is misleading because both the v2 and v3
services are present on that endpoint.
This is particularly confusing because we set this as the OS_AUTH_URL
endpoint and it makes it seem like we are doing v2 authentication when
we are not.
Change-Id: If73735026079fb19ca5bd44b3a4dc1f507b5c99d
With [1] glance_store introduced default settings for user_domain_id and
project_domain_id. Sadly since these are always passed to the keystone
client, they override any settings to user_domain_name and
project_domain_name that are made in the config, leading to authentication
failures.
So as a workaround until [2] is fixed, we explicitly place the corresponding
domain_ids into the config.
[1] https://review.openstack.org/297665
[2] https://bugs.launchpad.net/tempest/+bug/1620999
Change-Id: Ica81a1a176614392291f2db4cc6398ed30663aed
To support multinode testing where we just copy the CA to all the
instances don't remake the CA if it already exists.
The end result is that you can trusty a single chain and all your
clients will be happy regardless of which host they are talking to.
Change-Id: I90892e6828a59fa37af717361a2f1eed15a87ae4
TIL:
Similarly, all the END rules are merged, and executed when all the
input is exhausted (or when an exit statement is executed).
i.e. matching YUM_FAILED calls "exit", which falls through to the END
rules which calls "exit result" ... which is zero. i.e. if the return
code is 1 then we actually hide that and return with zero.
This is rather annoying because errors that should halt to alert us of
a package install failure pass through, only for you to have to debug
much later on seemingly unrelated problems.
This always sets "result" and thus should be returning the right
thing. I've updated the documentation to hopefully make it clearer
what's going on.
Change-Id: Ia15b7dc55efb8d3e3e945241b67a468b8a914672
In Debian jessie and later release,there is no packages
called "qemu-kvm" for AArch64. Also modify the libguestfs0
packages for AArch64
Closes-bug: #1612182
Change-Id: I5eb6bd137896eb9abfc4f8dbb41b41105e4820cd
Signed-off-by: Kevin Zhao <kevin.zhao@linaro.org>
Instead of only using the userrc_early when keystone
is an enabled service we will do it on all runs of
stack.sh. This way services can be split up more
across devstack nodes, and you can do configuration
requiring credentials on nodes that don't install
keystone.
Change-Id: I74574ae9f45a74bcbcc8e3149228ecb795ab4fb7
The flag ENABLE_DEBUG_LOG_LEVEL indicates if this should be
set or not.
This will now be supported in Neutron.
Change-Id: I3afe0546b379873247fee1ef9f4cc2708a7b5713
Some systems may have more than one default route.
Set up iptables NAT rules on all v4 default route devices.
Accept RAs on all v6 default route devices.
Closes-Bug: #1624773
Change-Id: If58509297497ea33c6c156f083a4394000bd0561
The use_usb_tablet option is replaced by the pointer_model
option.
Depends-On: Id18b5503799922e4096bde296a9e7bb4f2a994aa
Change-Id: Ic2a49f88df988c6404c1c72e9ee28a487e4f7908
In some initialization conditions (having never ran stack.sh) the
result of unstack.sh is dependent on if the user had previously
installed lvm2 or disabled the cinder service.
This change makes all results the same with a bit of LBYL.
There's also a drive-by to put a comment back where it belongs after
being accidentally moved in the related change.
Related-Change: I09b1a7bee0785e5e1bb7dc96158a654bd3f15c83
Change-Id: I9a7e052677d60cbbbdd582877f3c6c48c387f668
Closes-Bug: #1619195
Keystone had a problem where there was a memcached socket
opened very early on startup which then got shared between
worker processes when running under uwsgi. This can be
prevented by setting lazy-apps so this is the recommended
setting.
See http://uwsgi-docs.readthedocs.io/en/latest/ThingsToKnow.html
"""uWSGI tries to (ab)use the Copy On Write semantics of the
fork() call whenever possible. By default it will fork after
having loaded your applications to share as much of their
memory as possible. If this behavior is undesirable for some
reason, use the lazy-apps option. This will instruct uWSGI to
load the applications after each worker’s fork(). Beware as
there is an older options named lazy that is way more invasive
and highly discouraged (it is still here only for backward
compatibility) """
Change-Id: I6f271dc906528f0c86060452deaf15df81b267d2
Related-Bug: 1600394
Removing the explicit enablment of Neutron services, as with [1] they are configured as defaults in stackrc.
[1] https://review.openstack.org/#/c/350750/
Change-Id: Ic8910cd28fe37842f7d824e68bd2ea705e7e52de
memcache_pool is there to keep a limited number of thread-associated
connections open rather than a connection for every thread. If you
don't have a huge number of threads it doesn't offer anything.
Keystone is an example of a service where memcache_pool doesn't
improve things -- eventlet isn't supported anymore and more threads
is not useful due to GIL.
As such, keystone cache backend is changed to dogpile.cache.memcached.
See https://review.openstack.org/357407 for the oslo.cache help text
change.
Change-Id: I4452a8c4968073cdea4c0f384453a5a28519fa08
CEPH_CONF does not exist anymore, resulting both cinder-volume and
cinder-backup being configured with an empty rbd_ceph_conf option.
Using CEPH_CONF_FILE to fix this.
Change-Id: I1aa590aba900a4a94698917e45a0ea5c6f497f18
Signed-off-by: Sébastien Han <seb@redhat.com>
The plugin creates subnetpools but does not use them when creating the
default subnets. It uses CIDR values that overlap with the
default pools. Change this to use the subnetpools.
Change-Id: I6171c13507e420f146801d323cb1011be36c1e8c
Closes-bug: 1613717
Before the code in the extra.d plugins was removed from the devstack
tree they could define the order they ran. When this code is decomposed
into a plugin, there is still a need to do some form of ordering. This
caused problems with the Ironic devstack plugin and Tempest because the
code is run in this order:
1. The tempest configuration is run from extra.d, processing
DEFAULT_INSTANCE_TYPE, and writing the flavor_ref to tempest.conf
2. The Ironic devstack plugin is run, creating the flavor needed for
DEFAULT_INSTANCE_TYPE
This leads to build failures as tempest can not find the required
flavor, so it writes which ever flavor it can find at the time into
flavor_ref. Ironic now has code it its devstack plugin duplicated from
the tempest plugin to work around this problem until this is merged.
This patch fixes this by using the test-config phase to move the tempest
plugin as late as possible in the devstack process.
Change-Id: I3d98692e69d94756e0034c83a247e05d85177f02
Currentlly, the *.pyc files could not be removed in any scripts or
functions. But the redundant files would lead stack.sh not to find the
correct script for some versions after branch switched from master to
stable/mitaka in migration_helpers.sync_database_to_version.
So this commit adds the process of cleaning all the *.pyc files in
clean.sh.
It is needed to execute clean.sh before re-stack.sh to prevent the
exception.
Change-Id: I9ba0674d6b20b13c0a26b22cd5d1939daa121a94
Closes-Bug: #1599124
Add a VirtualHost that defines the necessary options for
enabling SSL. The existing keystone Apache configuration already
does all the location handling.
Change-Id: I836a471a7258f14f051d3dd8bdb428286b5a11aa
This will have devstack setup the Cinder internal tenant and generic
image-volume cache by default. If left alone it will use reasonable
defaults.
More information about configuration options and the cache can be found
here: http://docs.openstack.org/admin-guide/blockstorage_image_volume_cache.html
As part of this we switch the default lvm type to thin so it will
work more efficiently with the image cache.
Change-Id: I0b2cc261736f32d38d43c60254f0dc7225b24c01
Implements: blueprint cinder-image-volume-cache
2016-06-24 18:09:16 +00:00
318 changed files with 13984 additions and 11230 deletions
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Jeremy Stanley