Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: I39d085726551ac1db2531d474d9c407840f5ffc0
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
This prevents from exiting the job without pushing the logs
to artifacts if the functest loop fails.
JIRA: FUNCTEST-443
Change-Id: I6925b5855326df809dbce9f389b10be8a00cf8de
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
Change-Id: I07d7f0a3cdef3ebe1dec07156791f181a149ba19
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
|
|
This change contains updates that impact compass and joid jobs due
to updates in functest and yardstick that remove brahmaputra branch.
Please note that the triggers for colorado haven't been created yet.
It will be done once the basics in place.
Change-Id: Id5eb8c406d34ff9f3ac59f73243034cbf49b23f0
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Daily jobs for Colorado branch for installer and test projects
have not been created yet and it needs to be done via separate
patches.
Change-Id: I34517e89dfc502ce5741733e01bf8425d513df02
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
|
|
Change-Id: I61b68a7db60193f5525de7151e43b7752e5dd604
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
For all zte pods, the fuel ip is not 10.20.0.2 . It need to override.
Change-Id: Ib9533c28d9d03fdfc4804955de9eee7501453cd9
Signed-off-by: wu.zhihui <wu.zhihui1@zte.com.cn>
|
|
1. CI scenario: os-odl_l2-nofeature-ha using Fuel installer
2. Adjust qtip daily job from zte-pod1 to zte-pod2
Change-Id: I3e4e1adf3a6865a4cc9be81522472c781a8aea4f
Signed-off-by: Julien <zhang.jun3g@zte.com.cn>
|
|
Deployment in zte-pod3 successfully but yardstick and fucntest
can not be triggered.
JIRA: PHAROS-247
Change-Id: I5407f8c51bac83263b9f165558a38fa0db199868
Signed-off-by: Julien <zhang.jun3g@zte.com.cn>
|
|
JIRA: FUNCTEST-412
Change-Id: Ic23f5629e00d4adec62f6021bfd859f439e12825
Signed-off-by: SerenaFeng <feng.xiaowei@zte.com.cn>
|
|
Some installer/scenarios take too long to execute the
tests and the job times out.. Although this is not the
best solution, we can let the installer teams run the
tests to see why it takes too long sometimes)
Change-Id: I37cc393c65408741c5f0ecb9cc12ad6297c31d6d
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
Change-Id: Id06771db96d1ca014c153ffc2b7567d6b2ea07ea
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
Change-Id: I405bcbf26965a5290fb8dad9786881251e6aea81
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Change-Id: If63cca009f539125ce877fa05f40178bb1c71247
Signed-off-by: meimei <meimei@huawei.com>
|
|
This is the first step towards having alignment of the job structure, applying
commit gating and use of labels for all installer jobs.
Apart from the above, it also
- puts all CI baremetal and virtual PODs together
- gets rid of use of slaves directly for CI baremetal and virtual PODs
- switches labels for all CI baremetal and virtual PODs using below labels
{installer}-baremetal
{installer}-virtual
- puts all CI build slaves together and switches to labels - except huawei-build
- puts all none-CI PODs together
Please note that this change does not impact the overall job structure.
Change-Id: I2d8d15799a57fe7e063d39f81b96d19c218b100e
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
|
|
Something in Apex is pushing the RETURN rule down the chain, and adding
REJECT reject-with icmp-port-unreachable. This enhances the check to
make sure not only does the rule exist, but it is at the top of the
chain.
Change-Id: I527bc45c30008aaf30d5bc9e1d474a14065b3fdf
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
|
|
Change-Id: Ice55eece4aad6b9b8a1f51ebac13e26251b50708
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Change-Id: I2d1dfc4c9c09199d2ced56ce47f3376eb8ea8f81
Signed-off-by: Morgan Richomme <morgan.richomme@orange.com>
|
|
Change-Id: I4e2bdd499d60c07230d5f6a15a2c9fd9a9c8efdd
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
This is a required field for JJB > 1.4
Change-Id: Ifb6f5d4b3a28f2d85f546151ca9aca4a4061f3ef
Signed-off-by: Trevor Bramwell <tbramwell@linuxfoundation.org>
|
|
Change-Id: I2212c4e52b6ac7df017b6fda17df2b3a0e4a9406
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
Change-Id: Iea61baaca6f52fc794011aba40195c76c6da6b68
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
Change-Id: I53e7e208d380beb2799182dc85394215d7397f7c
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
This reverts commit ca27977f61fe104a5252cdc40b42c6c774724530.
Change-Id: I7eedfab682fc9f46c128c4164983ad9ae625b384
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
JIRA: APEX-122
Change-Id: I61e04f8e65fe7c35350cd5ab6d5190ffa9ed311f
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
JIRA: RELENG-116
Change-Id: I6aaa933616c6bb2d31ebcd51fbccaac6ed92aef4
Signed-off-by: Luke Hinds <lukehinds@gmail.com>
|
|
|
|
JIRA: RELENG-116
Change-Id: I178bc9046b3dc37b29b1b2fffd5c477ce4971afc
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
We reconfigured centos build server now and it is important to
know which jobs require centos and which ones ubuntu. The machines
with ubuntu were labelled with opnfv-build previously, preventing
us from keeping track of which projects require what OS. This tries
to solve that.
Change-Id: I1fb2912ec49f5bc2781853e500508d9992d59fbb
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
JIRA: FUNCTEST-275
At least until we fix the issue with the python logger showing
the output twice.
Change-Id: I121b6164cd9dc263b0c6f87afb661f853bcca708
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
Change-Id: Ia56ceefe5bf93cf15f89133c40f0e23c678e11d3
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
This change switches compass jobs to use labels instead of tying
jobs to slaves.
The changes in detail are:
- Compass jobs can roam on any pod that is labelled with compass-baremetal
and compass-virtual.
- Above means that there is no branch/pod allocation anymore.
- New jobs are created in order to get rid of pod names from them and just
use baremetal or virtual in names.
- Old jobs are removed.
- Related test jobs are updated.
- huawei-virtual label is renamed to compass-virtual since we still have
installer specific slaves, not company specific. In future the job
might end up on a none-Huawei slave too.
- huawei-pod2 is not a CI POD so it is left in none-CI PODs section.
- block relations and triggers might require fine tuning and left to compass
project to decide how to adjust
Job structure remains same; one parent job calling 3 downstream jobs:
deploy, functest, yardstick.
The OS to deploy (trusty or centos7) needs to be solved in a separate patch
in a sustainable way.
Change-Id: I22bfaff36640667b6764f0e303a3286af0eb41b2
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
This change switches joid jobs to use labels instead of tying
jobs to slaves.
The changes in detail are:
- Joid jobs can roam on any pod that is labelled with joid-baremetal
and joid-virtual.
- Above means that there is no branch/pod allocation anymore.
- New jobs are created in order to get rid of pod names from them and
just use baremetal or virtual in names.
- Old jobs are removed.
- Related test jobs are updated.
- POD specific parameters are updated and instead of adding pod
specific parameters, they now become parameters for all the pods
labelled with same label. It is not a problem at the moment as
both Intel PODs (pod5, pod6) have same parameters but a different
pod with different parameters will require alignment between pods.
Job structure remains same; one parent job calling 3 downstream jobs:
deploy, functest, yardstick.
Change-Id: I30d270cd05ba3756f6f161d65ee5160eab6b80f4
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
This change switches fuel jobs to use labels instead of tying
jobs to slaves.
The possibilities and changes introduced with this approach are
- If things do not work as expected, slave-label parameter can
be set to certain slave name to get back to old way of
tying jobs to slaves.
- If a scenario requires a different slave, it can still be
created using baremetal or virtual but the label can be
different, pointing to a slave that is capable of running
that scenario. (fuel-baremetal-multisite, etc.)
- The auto-trigger-name for brahmaputra is moved under the
branch anchor and can be used like this if triggers for all
all the jobs for a certain branch need not to be triggered
automatically but job is still kept enabled in case if some
crucial bug needs to be tested.
- No of max concurrent builds stepped to 4 in order to use
all the labelled slaves available. This will require some
adjustments before colorado jobs start to run in order to
share resources between branches fairly.
- All the fuel CI PODs except lf-pod1 have been removed.
Change-Id: I91eca6743bb3d34263e634f9b74550efa1e2ddbe
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
JIRA: COMPASS-393
Change-Id: Ib079cb543f7ed9ae932ea3af99a8c3f187dd0cbc
Signed-off-by: meimei <meimei@huawei.com>
|
|
Change-Id: I22bdffac5888656ee95c458d42c0f6621c822f7f
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Change-Id: I546b1b323db2d1839335a34e8293b27f368b0626
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Change-Id: If58c84e6f431db88eed3376e446479826da71502
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Slave names will be as below.
- huawei-us-deploy-bare-1 -> huawei-pod1
- huawei-us-deploy-bare-2 -> huawei-pod2
- huawei-sh-test-2-defaults -> huawei-pod3
- huawei-sh-pod6 -> huawei-pod4
- huawei-us-deploy-vm-1 -> huawei-virtual1
- huawei-us-deploy-vm-2 -> huawei-virtual2
- huawei-us-deploy-vm-3 -> huawei-virtual3
- huawei-us-deploy-vm-4 -> huawei-virtual4
huawei-sh-pod4 will be removed from Jenkins.
Please note that this change will result in removal of existing jobs.
This is not that crucial since the use of labels will result in same.
Change-Id: I63798c799a402b1694ee0767ffdb45a597897cd9
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Change-Id: Ib01658ed1e8c85b6408d32a1b367702bc6b32abd
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
orange-test1 -> orange-pod5
orange-fr-pod2 -> orange-pod2
Change-Id: Id5a31db98d349bef654fb59ab03e52e6cc50611f
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
JIRA:RELENG-87
Change-Id: I4ca1ea6a206fcadc94f813e33bae4751236ee24d
Signed-off-by: zhifeng.jiang <jiang.zhifeng@zte.com.cn>
|
|
That output is irrelevant and doesn't help to debug anything, and
it is very extensive.
Change-Id: If97ff388d5ab53c67d7cd5c5e6f2fff1f202e6a4
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
This makes it possible to pass extra environments or volumes to docker
by listing them in a file at ${HOME}/opnfv/functest/custom/params_${DOCKER_TAG}.
This way different parameter (env vars and volumes) can be specified for stable
and master.
Change-Id: I010fae9802dd413f0ca6ae9fb3ef0f03b81db3c2
Signed-off-by: Ciprian Barbu <ciprian.barbu@enea.com>
|
|
|
|
Change-Id: If79d4a457712465af1e4be1aca8a51ad270d65fa
Signed-off-by: Morgan Richomme <morgan.richomme@orange.com>
|