Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
|
|
Functest/yardstick jobs are missing for Gambia. This patch
adds them.
Change-Id: If65bc5927dba745ff5dbd858b4e612c0f48e1ab1
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
|
|
|
|
When we promote artifacts we do not want to include the apex branch (for
example "gambia") in the gs path URL for where snapshots are uploaded.
We may use a different apex branch to promote older OS versions of
snapshots, but we still want a unified url of just
apex/<os_version>/<topology> for our snapshot artifacts.
Change-Id: I7dfe40558ce05d04e1805884cd6a47c82220581e
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Add fuel CI PODs and armband CI PODs for gambia stream in yardstick-daily jobs
Change-Id: I1b685ca4859eb476f56ceddb4368215605ad7ae5
Signed-off-by: Cristina Pauna <cristina.pauna@enea.com>
|
|
This patch should be merged after https://gerrit.opnfv.org/gerrit/#/c/63305/
Change-Id: Ic809d8de200ba1b55e18c3079587404e4077cea0
Signed-off-by: xudan <xudan16@huawei.com>
|
|
|
|
Update envs only once with IMAGE_PROPERTIES and FLAVOR_EXTRA_SPECS
if installer type is fuel or deploy scenario is ovs or fdio
Change-Id: I7c1e997cf95d30972688e697574458a3a0f1e55c
Signed-off-by: Delia Popescu <delia.popescu@enea.com>
|
|
We want the ability to be able to force a promotion and skip CSIT. This
patch introduces a variable we can use when manually triggering a
promotion job to force promotion. The use case here is for HA scenarios
we have no snapshots because CSIT always fails. We want to be able to
force promotions for this case to at least have some snapshot available.
Change-Id: I7e81b458f4169596929f032e9394695477fbb940
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
This brings in:
Skip vote for releng-rtd from gerrit trigger
and three changes after it that are either innocuous or important.
xmldiff:
https://hastebin.com/zutufehafa.diff
Change-Id: Ic5c67ff5d6aebc6fc2149b212acc3bcf7cf9951b
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
|
|
Change-Id: Id789121634fc1a009878b61bcf4756ee4009c99d
Signed-off-by: wutianwei <wutianwei1@huawei.com>
|
|
|
|
|
|
|
|
|
|
Moves the jjb-version configuration from globals to releng CI jobs. This
was not being pulled into the '{project-name}-ci-jobs' configuration and
kept our JJB version at 2.2.1.
Change-Id: I33a62e639bbca6d519957515af591615543d27b7
Signed-off-by: Trevor Bramwell <tbramwell@linuxfoundation.org>
|
|
apex-virtual-{stream} is not being blocked and running at the same time
while csit tests are running. This should stop it from running until the
promotion job is finished.
Change-Id: Ie164bbee08253a7e79892d36780b66fe0ad2cf97
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
|
|
|
|
|
|
|
|
We need the fix:
'Allow running only on one of nodes with label parameter'
https://review.openstack.org/#/c/590356/
This fix is present in 2.5.0
Change-Id: Iff9e9e6357573db382d94b688b7cfe7d42d4023c
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
|
|
Has been offline for > 1 week
https://build.opnfv.org/ci/job/functest-core-docker-build-arm64-gambia/
is showing
There are no nodes with the label ‘Job triggered without a valid online
node, given where: arm-build6’)
Change-Id: Ia00281f9c05fac9b16bd914063ccffeca3fb135b
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
|
|
With changes to deploy scenario files for gambia, we still need the
logic to pull in the right csit-env. That logic broke as it depended on
DEPLOY_SCENARIO with the changes to scenario names. This patch fixes it
to check openstack version (OS_VERSION) instead.
Change-Id: Ie130efd0fc3cb7f09a3fa4e896ad461d24394777
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
In Apex, some random jobs are queueing to intel-pod11 which does not
exist anymore. The problem is, when these jobs queue they are blocking
apex daily functest, as it looks like the job is only set to allow 1 job
globally at a time. This change bumps the max concurrent global jobs to
10.
Change-Id: I62d458a92abfa35bb0d1c459ccf4ffedfd51e66d
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Add arm-virtual3 pod into CI pod list for running
the functest-kubernetes cases on arm platform.
Change-Id: Iae1881320956795304896d8ea2c1790430f919da
Signed-off-by: Jingzhao Ni <jingzhao.ni@arm.com>
|
|
|
|
|
|
Change-Id: I36b2f245aa2044ebcdab5122d3b30ae1d9fb1788
Signed-off-by: wutianwei <wutianwei1@huawei.com>
|
|
|
|
Since CSIT does not pass on HA, but we still want to promote some
snapshots for HA, reduce the CSIT suites to L2 and L3 only.
Change-Id: Iaf739f972211c1e94de781a006065d7130ddf1c3
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Now that Gambia branch is used for queens promotion, scenario files use
nofeature instead of using branch name.
Change-Id: I76e8f851bc6ef337301d1367dcafd798b212bb38
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Change-Id: I2c9c1f042b4b13af58526e23a44680304313cee9
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
|
|
|
|
A change to the NodeLabelParameter config in JJB caused any jobs using
the plugin to possibly have their jobs triggered on builders which
weren't online.
This in turn caused the queue to backup and newer jobs are blocked
waiting for older ones to complete, which never happens as they're never
queued on an online builder.
The 'all-nodes' parameter ensure jobs that are set to run concurrently
don't trigger across all available nodes at once, as this will include
offline nodes as well. As this is the default of the plugin, projects
which need it can enable it individually.
Change-Id: Ia690eef078209b6b056dad85613dda6868b18271
Signed-off-by: Trevor Bramwell <tbramwell@linuxfoundation.org>
|
|
|
|
Missed a check where we install packages based on old package
convention.
Change-Id: If1da8db9b10b81ea5825b7548509be29e08e6f91
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Change-Id: Ibe83eb4131d8ff553aa81e343cfe5ae8c19b4cf6
Signed-off-by: Ryota MIBU <r-mibu@cq.jp.nec.com>
|
|
When checking files, it shouldn't exit when getting an non-zero results.
However, the scripts set -e at the begining.
In order to continue the following steps when the file doesn't exists, replace
all "return 0" or "return 1" to be "echo 0/1".
Change-Id: I28ea81d52182d4dc5db56d47abbfbc1761f17d81
Signed-off-by: xudan <xudan16@huawei.com>
|
|
Change-Id: I4cecaf7302fa1f39aa1e8b2c3e34f11ee49b4359
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
Global JJB now has non-JJB YAML configuration and requires action on the
ci-management repo when upgrading to this version of Global JJB to
prevent JJB from picking up these YAMLs as config. Follow the
instructions below BEFORE upgrading globall-jjb:
cd <git-root>
git mv jjb/global-jjb global-jjb
mkdir jjb/global-jjb
ln -s ../../global-jjb/shell jjb/global-jjb/shell
ln -s ../../global-jjb/jjb jjb/global-jjb/jjb
git add jjb/global-jjb
git commit -sm "Prepare repo for global-jjb v0.25.0"
Git diff confirms no changes.
diff -bur . /tmp/releng/job_output/
Change-Id: I6d7239b8bab8ef08368f833d9a64cd94a824b158
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
|
|
From the release notes:
Upgrade to global-jjb v0.24.6 before performing this upgrade. This
ensures that jjb-verify job pulls in a regex fix that will allow it to
verify the v0.25.0 upgrade.
Change-Id: I48335de599b46b7e6c70f8bb77fb8c254cfe85f4
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
|
|
|
|
|
|
Change-Id: I89040a4f7ac493efe8a11a268ce891c156e53c18
Signed-off-by: jenkins-ci <jenkins-opnfv-ci@opnfv.org>
|
|
jjb/lf-ci-jobs.yaml has jjb-version: 2.2.1 as a minimum
change block-level to blocking-level:
https://review.openstack.org/#/c/608092/
We will need to change this back once we move past
jenkins-job-builder>2.5.0
Change-Id: I12d0f55ee3c2dad20340ebf0fa8428ee1ad8484c
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
|
|
Change-Id: I388d5b46d206655990f3fb2723a90e8ed4397c89
Signed-off-by: Emma Foley <emma.l.foley@intel.com>
|
|
|