162 Commits

Author SHA1 Message Date
Riccardo Pittau
8377e1a06c Rename ironic-tox-unit-with-driver-libs-python3
Rename ironic-tox-unit-with-driver-libs-python3 and adjust tox
accordingly.

Change-Id: I3522f4a1c084c9cc2f7565f16ec0c790a1767377
2020-03-10 10:14:42 +01:00
Iury Gregory Melo Ferreira
2293c7dfad Set abstract for ironic-base
In Zuulv3 the abstract attribute is used to indicate that the job is not
intended to be run directly[1]. This is the case for our ironic-base
job.

[1] https://zuul-ci.org/docs/zuul/reference/job_def.html#attr-job.abstract

Change-Id: I128be871e45de9f5f27338117d58a63889eca615
2020-03-03 14:50:59 +01:00
Zuul
361a97584f Merge "Revert "Move ironic-standalone to non-voting"" 2020-02-14 23:25:01 +00:00
Julia Kreger
263c3c6073 Revert "Move ironic-standalone to non-voting"
This reverts commit e1e89f8831fd1b28b7ee373fc9504ce4eaf2c4c5.

The commit was a temporary change to allow us to unblock
the ironic CI gate as we had cross-repository fixes that
needed to land before we would be able to have a passing
and stable ironic-standalone job.

Change-Id: Ib1ed0a35955ab7c53b814048887e725234699d9a
2020-02-14 18:18:54 +00:00
Zuul
0a8b62972a Merge "Lower tempest concurrency" 2020-02-14 00:31:29 +00:00
Iury Gregory Melo Ferreira
e1e89f8831 Move ironic-standalone to non-voting
Let's keep standalone non-voting till we can stabilize the
failures on it.

Change-Id: I5c0d23f8cf9f377f013a96bfe30189df25f9c7c5
2020-02-13 09:35:32 +01:00
Iury Gregory Melo Ferreira
caa73a1101 Lower tempest concurrency
Let's only run 2 concurrency tempest tests to see if we can
decrease the failure ratio for ironic-standlone job

Change-Id: Ifb6258df26236cf376742a051602321e9b67a52c
2020-02-12 22:31:10 +01:00
Riccardo Pittau
20142434b6 Use bionic job for bifrost integration
The bifrost bionic job is alive again so we can change it here.

Change-Id: I6ee566d111f4d7aa8df12b13c11b8810bf078d25
2020-01-21 19:43:29 +01:00
Dmitry Tantsur
8f2ddc0b21 Lower RAM for DIB jobs to 2 GiB
The CentOS 8 images we use now are much smaller than the previously
used CentOS 7 images, so hopefully we can require less RAM.

Change-Id: I39fc7c202c1f148e28c22739f3163081bfa7e0a0
2020-01-14 00:30:03 +01:00
Zuul
9669948732 Merge "devstack: switch to using CentOS 8 DIB ramdisks by default" 2020-01-09 12:32:24 +00:00
Dmitry Tantsur
8c3ab7de0f CI: make the metalsmith job voting and gating
This job is the only job that is currently testing partition images
with local boot. Due to its usage of a custom centos instance image,
there is some potential instability, but recent runs [1] have been
reliably green.

[1] https://zuul.opendev.org/t/openstack/builds?job_name=metalsmith-integration-glance-localboot-centos7

Change-Id: Iefbf94754da58c130eccba6fa24efac684ee4b41
2020-01-08 19:22:17 +01:00
Zuul
b39ca69061 Merge "Explicitly use ipxe as boot interface for iPXE testing" 2020-01-07 16:09:04 +00:00
Zuul
b886e8b196 Merge "Increasing BUILD_TIMEOUT value for multinode job" 2020-01-07 13:41:49 +00:00
Dmitry Tantsur
f110afc4f9 devstack: switch to using CentOS 8 DIB ramdisks by default
The CentOS 7 ones no longer work because of the Python 3 transition.

Change-Id: I68fc6f87879a53198810ad64f9570c41ccf8ce7f
2020-01-03 16:28:54 +01:00
Kaifeng Wang
697048b47e Explicitly use ipxe as boot interface for iPXE testing
We are about to drop pxe + ipxe_enabled=True combination for iPXE
testing, so let's make it explicit in the jobs.

Change-Id: I0e99e6336aa47857c06ba2f6869e66580159a9be
Story: 2007003
Task: 37779
2020-01-02 14:00:42 +08:00
Zuul
67f5a6bf5c Merge "Disable automated clean on newer jobs" 2019-12-19 19:56:45 +00:00
Zuul
cd30ef1374 Merge "Extend service timeout" 2019-12-19 06:54:13 +00:00
Zuul
8d8bfc9cc5 Merge "Tune down multinode concurrency" 2019-12-18 15:47:35 +00:00
Julia Kreger
d0d221361d Tell the multinode subnode and grenade to use /opt
It was changed on the primary node, but not the subnode
of the standard mutlinode job, and not updated on our
grenade job.

Changes subnode and grenade configuration to use it.

Change-Id: I8f67e3e3423e72944d22d65b788af5792ee900f9
2019-12-17 06:10:35 -08:00
Julia Kreger
7b4e10c062 Disable automated clean on newer jobs
We simply do not need to run automated cleaning on most jobs
and it has been left enabled on a number of newer jobs.

Disabling it improves gate performance by reducing the overall
wait to start tests and cycle through test nodes as sometimes
spike in testing infrastucture utilization can create increased
latency which may cause cleaning to time out.

Change-Id: I2d0d33a5b62fed536785915b2a80d7af8b6f3e0d
2019-12-16 21:22:58 -08:00
Julia Kreger
f4d2a3127a Extend service timeout
Glance seems to be having issues getting started in time.

At least a couple times recently devstack exited because glance
was not running with-in 60 seconds. Logs in one case seemed to
lean to well over a minute, as such lets extend the timeout
and see if that stabilizes CI further.

[0] https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_b45/697584/10/check/ironic-tempest-ipa-wholedisk-bios-agent_ipmitool-tinyipa-indirect/b456a36/

Change-Id: I7d4457a456dedead4ad0d6a444be23a901b42111
2019-12-16 21:04:01 -08:00
Julia Kreger
da704d9ec8 Tune down multinode concurrency
Reduce the multinode concurrency and number
of VMs so we don't run Rackspace VMs out of
disk space.

This is an issue that was reported by the
infra team.

[0]: http://eavesdrop.openstack.org/irclogs/%23openstack-ironic/%23openstack-ironic.2019-12-16.log.html#t2019-12-16T17:38:15

Change-Id: I867bac9e247605e561cc7ee4522f70ce6de368d4
2019-12-16 11:56:04 -08:00
Dmitry Tantsur
7a7e9689a3 Implement managed in-band inspection boot for redfish-virtual-media
The current vmedia job is switched to test inspection as well.

Change-Id: I077d9dd4b527fc090833587e62af352b20788458
Story: #1528920
Task: #37646
2019-12-13 00:32:38 +00:00
Ilya Etingof
200959b903 Add a CI job to UEFI boot over Redfish virtual media
Adds a job to boot a node over Redfish virtual media in UEFI boot
mode.

Turns similar BIOS boot mode job non-voting to avoid CI
failure on non-essential job failure.

Change-Id: Iaabd09a7acbbd692dd72adcf6867981bcc30877e
2019-12-13 00:31:58 +00:00
Zuul
746c988663 Merge "Add a CI job to legacy boot over Redfish virtual media" 2019-12-12 18:43:14 +00:00
Riccardo Pittau
bd5a12f753 Increasing BUILD_TIMEOUT value for multinode job
After some testing, it seems that the real cuplrit for all the
failures we're seeing is that we're not giving enough time for the
job to complete.
This patch increase the build timeout value to 2400 secs, which
should be enough time for the job to exit successfully.

Change-Id: Ia93d7ac26a004849b97aeb98d5c5cff3df14da0f
2019-12-12 14:50:40 +01:00
Dmitry Tantsur
570317e400 CI: do not enable rescue on indirect jobs
They get it enabled because of their parent.

Change-Id: I12083867fa5e0020410e87fb06bc64ef42bc4d38
2019-12-10 17:11:27 +01:00
Ilya Etingof
dcd701b4a0 Add a CI job to legacy boot over Redfish virtual media
Adds a job to boot a node over Redfish virtual media in legacy
BIOS boot mode. For sake of unification, similar PXE job has been
renamed.

Apparently, libvirt-based Redfish BMC emulation requires node
power-cycle for libvirt domain configuration changes to take
effect.

Change-Id: I2c7fc26b082db6eeae1abcd67362884a9c72c109
2019-12-04 14:59:08 +01:00
Zuul
388cace718 Merge "Add an ironic-inspector job with managed boot" 2019-12-03 20:25:11 +00:00
Ilya Etingof
3eb1ca6338 Make redfish CI jobs pulling sushy-tools from git
Otherwise Depends-On sushy-tools does not seem to work.

Depends-On: https://review.opendev.org/693561
Change-Id: Iea32b98a57713b70b535eb524b165cbeb141199f
2019-11-28 12:22:56 +01:00
Dmitry Tantsur
71c03410ae Add an ironic-inspector job with managed boot
Change-Id: Iacc26bcebb16b8baec5cf03cf53b5e7a22ea0ea7
Depends-On: https://review.opendev.org/#/c/688920/
2019-11-27 16:28:26 +01:00
Iury Gregory Melo Ferreira
b869b46c63 Drop python 2.7 support and testing
OpenStack is dropping the py2.7 support in ussuri cycle.

Ironic is ready with python 3 and ok to drop the
python 2.7 support.

Complete discussion & schedule can be found in
- http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010142.html
- https://etherpad.openstack.org/p/drop-python2-support

Ussuri Communtiy-wide goal - https://review.opendev.org/#/c/691178/

Change-Id: I3cc36d79ffabf31607d503dada9e7512c639647c
2019-11-22 14:06:45 +01:00
Zuul
630c85126b Merge "Change integration jobs to run under Python3" 2019-10-26 08:07:42 +00:00
Zuul
d471beed61 Merge "CI: add ironic-python-agent-builder to the multinode job" 2019-10-24 14:06:18 +00:00
Iury Gregory Melo Ferreira
bba990c7e0 Change integration jobs to run under Python3
This patch enables python3 by default in the `ironic-base` job.
All jobs that have `ironic-base` as `parent` will be running under
python3 by default.

- Switch parent/child for `ironic-tempest-functional-python2` and
`ironic-tempest-functional-python3`

- Renamed `ironic-tempest-ipa-partition-pxe_ipmitool-tinyipa-python3`
to `ironic-tempest-ipa-partition-pxe_ipmitool-tinyipa` since it was
running under python3 that it's the default now

Depends-On: https://review.opendev.org/690537
Change-Id: Ibc503baa572c2746e1e703653ddf040bad5e8363
2019-10-24 10:20:52 +02:00
Zuul
06bf5df3be Merge "CI: limit rescue testing to only two jobs" 2019-10-23 17:06:02 +00:00
Dmitry Tantsur
a38b1c8340 CI: add ironic-python-agent-builder to the multinode job
It is not unconditionally required by the devstack plugin.

Change-Id: I1d60323829e6bbc8e835314b2eef8daa7f353631
2019-10-23 18:19:15 +02:00
Dmitry Tantsur
a2472c1518 CI: limit rescue testing to only two jobs
This feature is time-consuming and prone to race conditions. This
changes switches it off by default, leaving it on in two jobs:
the one running on Nova and the standalone one.

Change-Id: I7c50cb12065da05dad4ece0918883748ee4617e0
2019-10-22 17:20:51 +02:00
Iury Gregory Melo Ferreira
3db445ab5d Switch to ussuri job
Change-Id: I49082656880dbfcdb54ad0d2808e57eaa4070a56
2019-10-21 11:55:52 +02:00
Zuul
270efc1443 Merge "Change MTU logic to allow for lower MTUs automatically" 2019-10-18 02:35:07 +00:00
Julia Kreger
cf074202e5 Change MTU logic to allow for lower MTUs automatically
Change-Id: I40d357e716b589585493824cfd963fd5f579bd81
2019-10-11 10:07:50 -07:00
Iury Gregory Melo Ferreira
8cc6dbdd89 ironic-tempest-functional-python3 unused variables
The difference between the jobs `ironic-tempest-functional-python2`
and `ironic-tempest-functional-python3` should be only the python
version in use (for deployment and to run the functional tests).

Removing the variables `IRONIC_TEMPEST_WHOLE_DISK_IMAGE` and
`IRONIC_VM_EPHEMERAL_DISK`doesn't matter for the functional tests,
and they are set with different values in the python2 job.

Change-Id: I49ff241793f42de0f0cc9d98ae1f99940e581dcb
2019-10-09 11:00:01 +02:00
Dmitry Tantsur
78e763c61b Add a CI job with a DIB-built ramdisk
Also updates the devstack plugin to support pre-built DIB images
and removes the explicit mention of dhcp-all-interfaces element,
which is already a dependency of ironic-python-agent-ramdisk.

Change-Id: I564016528e80c738d6004c47f8df502a7748b311
2019-09-30 20:14:19 +02:00
Zuul
4b56fe29bd Merge "Make multinode jobs non-voting" 2019-09-21 04:19:11 +00:00
Zuul
5342ca17c0 Merge "Lower MTU override" 2019-09-20 23:21:11 +00:00
Julia Kreger
26c4a00c74 Make multinode jobs non-voting
As Dmitry suggested, we should likely just make this
a standard practice as end of cycle gate instability
makes multinode jobs blocking in nature.

We will revert this patch once things are happier,
hopefully sooner rather than later.

Change-Id: I17b853aabd1a2cddf64d73b3778d315092d5612a
2019-09-20 07:45:21 -07:00
Julia Kreger
1745eea83e Lower MTU override
For some unknown reason, new networks being created in the
fortnebula test VMs are defaulting to a MTU size of 1350 bytes.

This is based upon the physnet MTU, which does appear to be
lower on the fortnebula cloud.

Change-Id: I113973ba2a0a33c68f485d02af1ce8e12a6fbba4
Story: 2006589
Task: 36716
2019-09-19 17:24:09 -07:00
Dmitry Tantsur
625c71d48d CI: move libvirt images to /opt for standalone and multinode jobs
Moves the volume pool from / to /opt for the standalone and multinode
jobs, since on smaller nodes /opt is mounted from a larger ephemeral disk.

Change-Id: Idb4e235fd40ecd9829c3ad8a9dd1224f1c29a12e
2019-09-16 18:57:22 +02:00
Zuul
5df2ecbdf5 Merge "CI: move the fast-track job to the experimental pipeline" 2019-09-14 20:57:01 +00:00
Dmitry Tantsur
cfc0f5dac9 Combined gate fixes
We seem to be running out of disk space in standalone and mutlinode
jobs. Since we're creating a lot of virtual disks here, reduce each
to 4 GiB. No jobs need so much anyway. Reducing further will likely
hit issues with IPA assumptions of root devices.

Then correct setting OS_TEST_TIMEOUT in the multinode job. DevStack
variables do not affect tox runs, we need to use a new tempest
role variable instead.

Change-Id: Iceebf246b000cf77cbbf62e4edb9b06ee9dd7b71
2019-09-11 11:39:52 +02:00