- Infra - oVirt List Archives

[JIRA] (OVIRT-1743) ci please build and send to OST
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1743?page=com.atlassian.jira.… ] Barak Korren updated OVIRT-1743: -------------------------------- Epic Link: (was: OVIRT-400) > ci please build and send to OST > ------------------------------- > > Key: OVIRT-1743 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1743 > Project: oVirt - virtualization made easy > Issue Type: New Feature > Components: OST Manual job, Standard CI (Freestyle) > Reporter: danken > Assignee: infra > Labels: change-queue > > Can you please automate the frequent process, which is slow and error prone? > I am writing a patch, and would like to test it in ost. now I need to > * "ci please build" > * copy el7 url > * wait for el7 build to finish > * copy artifacts url to ovirt--system-test-manual, start it > * copy the ost run URL to the gerrit patch, so I have it for reference of failure/success. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1743) ci please build and send to OST
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1743?page=com.atlassian.jira.… ] Barak Korren updated OVIRT-1743: -------------------------------- Epic Link: (was: OVIRT-400) > ci please build and send to OST > ------------------------------- > > Key: OVIRT-1743 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1743 > Project: oVirt - virtualization made easy > Issue Type: New Feature > Components: OST Manual job, Standard CI (Freestyle) > Reporter: danken > Assignee: infra > Labels: change-queue > > Can you please automate the frequent process, which is slow and error prone? > I am writing a patch, and would like to test it in ost. now I need to > * "ci please build" > * copy el7 url > * wait for el7 build to finish > * copy artifacts url to ovirt--system-test-manual, start it > * copy the ost run URL to the gerrit patch, so I have it for reference of failure/success. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1759) Gerrit patches can reach CQ out-of-order
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1759?page=com.atlassian.jira.… ] Barak Korren updated OVIRT-1759: -------------------------------- Epic Link: (was: OVIRT-400) > Gerrit patches can reach CQ out-of-order > ---------------------------------------- > > Key: OVIRT-1759 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1759 > Project: oVirt - virtualization made easy > Issue Type: Bug > Components: Change Queue, Gerrit/git > Reporter: Barak Korren > Assignee: infra > Priority: High > Labels: change-queue, gerrit > > Is seems that despite the fact that all the jobs involved were designed to preserve proper patch ordering, there can be cases where patches trigger Jenkins jobs in the wrong order, which in turn end up putting the patches in the wrong order into the CQ. This in turn, causes the CQ to fail accurately detecting the patches that cause issues. > Here is en example of such an occurance. The following patches were merged in the order specified below: > # https://gerrit.ovirt.org/c/83917 > # https://gerrit.ovirt.org/c/83918 > # https://gerrit.ovirt.org/c/83919 > # https://gerrit.ovirt.org/c/83920 > The standard-enqueue runs putting them in the CQ however, seem to have run in a different order: > # http://jenkins.ovirt.org/job/standard-enqueue/6712/ (83917) > # http://jenkins.ovirt.org/job/standard-enqueue/6713/ (83919) > # http://jenkins.ovirt.org/job/standard-enqueue/6714/ (83920) > # http://jenkins.ovirt.org/job/standard-enqueue/6715/ (83918) > This resulted in CQ 'add' command running in the wrong order: > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12200/ (add 83917) > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12201/ (add 83919) > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12202/ (add 83920) > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12203/ (add 83918) > We need to come up with a way to make the system properly preserver order. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1759) Gerrit patches can reach CQ out-of-order
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1759?page=com.atlassian.jira.… ] Barak Korren updated OVIRT-1759: -------------------------------- Epic Link: (was: OVIRT-400) > Gerrit patches can reach CQ out-of-order > ---------------------------------------- > > Key: OVIRT-1759 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1759 > Project: oVirt - virtualization made easy > Issue Type: Bug > Components: Change Queue, Gerrit/git > Reporter: Barak Korren > Assignee: infra > Priority: High > Labels: change-queue, gerrit > > Is seems that despite the fact that all the jobs involved were designed to preserve proper patch ordering, there can be cases where patches trigger Jenkins jobs in the wrong order, which in turn end up putting the patches in the wrong order into the CQ. This in turn, causes the CQ to fail accurately detecting the patches that cause issues. > Here is en example of such an occurance. The following patches were merged in the order specified below: > # https://gerrit.ovirt.org/c/83917 > # https://gerrit.ovirt.org/c/83918 > # https://gerrit.ovirt.org/c/83919 > # https://gerrit.ovirt.org/c/83920 > The standard-enqueue runs putting them in the CQ however, seem to have run in a different order: > # http://jenkins.ovirt.org/job/standard-enqueue/6712/ (83917) > # http://jenkins.ovirt.org/job/standard-enqueue/6713/ (83919) > # http://jenkins.ovirt.org/job/standard-enqueue/6714/ (83920) > # http://jenkins.ovirt.org/job/standard-enqueue/6715/ (83918) > This resulted in CQ 'add' command running in the wrong order: > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12200/ (add 83917) > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12201/ (add 83919) > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12202/ (add 83920) > # http://jenkins.ovirt.org/job/ovirt-master_change-queue/12203/ (add 83918) > We need to come up with a way to make the system properly preserver order. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1774) Upstream source collector can fail to push on multi-branch projects
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1774?page=com.atlassian.jira.… ] Barak Korren updated OVIRT-1774: -------------------------------- Epic Link: (was: OVIRT-400) > Upstream source collector can fail to push on multi-branch projects > ------------------------------------------------------------------- > > Key: OVIRT-1774 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1774 > Project: oVirt - virtualization made easy > Issue Type: Bug > Components: pusher.py > Reporter: Barak Korren > Assignee: infra > Labels: poll-upstream-sources, upstream-source-collector > > When the upstream source collection code looks for similar patches in order to avoid pushing a new patch, it lookes for patches that contain the same change to the '{{upstream-sources.yaml}}' file, regardless of the branch they may belong to. > The code currently ignores the possibility that similar changes might be required for different branches, in which case different patches may need to be pushed. > The way of detecting similar patches should be changed so that either the branch name is included in the checksum that are used to identify a patch, or the query is limited to patches of the branch being handled. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1774) Upstream source collector can fail to push on multi-branch projects
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1774?page=com.atlassian.jira.… ] Barak Korren updated OVIRT-1774: -------------------------------- Epic Link: (was: OVIRT-400) > Upstream source collector can fail to push on multi-branch projects > ------------------------------------------------------------------- > > Key: OVIRT-1774 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1774 > Project: oVirt - virtualization made easy > Issue Type: Bug > Components: pusher.py > Reporter: Barak Korren > Assignee: infra > Labels: poll-upstream-sources, upstream-source-collector > > When the upstream source collection code looks for similar patches in order to avoid pushing a new patch, it lookes for patches that contain the same change to the '{{upstream-sources.yaml}}' file, regardless of the branch they may belong to. > The code currently ignores the possibility that similar changes might be required for different branches, in which case different patches may need to be pushed. > The way of detecting similar patches should be changed so that either the branch name is included in the checksum that are used to identify a patch, or the query is limited to patches of the branch being handled. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1782) Make sure all VDSM build dependencies for FC27 are met.
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1782?page=com.atlassian.jira.… ] Barak Korren updated OVIRT-1782: -------------------------------- Resolution: Won't Fix Status: Done (was: To Do) VDSM development had already moved on to FC28 > Make sure all VDSM build dependencies for FC27 are met. > ------------------------------------------------------- > > Key: OVIRT-1782 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1782 > Project: oVirt - virtualization made easy > Issue Type: New Feature > Components: CI client projects > Reporter: Barak Korren > Assignee: infra > Labels: vdsm > > There had been attempts recently to make VDSM build on FC27. > Those attempts were met with mixed success as issues kept surfacing. > Fro now all the FC27 jobs for VDSM had been disabled. > This ticket is for racking work of debugging issues to re-enable the jobs -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1831) Make `global-setup.sh` check if packages are installed before trying to install them
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1831?page=com.atlassian.jira.… ] Barak Korren reassigned OVIRT-1831: ----------------------------------- Assignee: (was: infra) > Make `global-setup.sh` check if packages are installed before trying to install them > ------------------------------------------------------------------------------------ > > Key: OVIRT-1831 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1831 > Project: oVirt - virtualization made easy > Issue Type: Improvement > Components: Jenkins Slaves > Reporter: Barak Korren > > The way '{{global-setup.sh}}' works currently is that it tries to install all the packages it needs in a single yum command and then runs in a loop and checks if the packages were installed properly. > If the packages are already installed this takes a negligible amount of time on CentOS slaves, but on Fedora it seems to take much longer and as much as 2 minutes. > We should try to make '{{global-setup.sh}}' faster by making it not try to install packages before checking if they are there, but instead check first and only try to install the missing ones. > Note that if we'll install multiple packages in a single command, we'll still need to verify they were properly installed after the yum/dnf command finishes. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1836) Optimize options normalization methods in stdci_dsl
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1836?page=com.atlassian.jira.… ] Barak Korren reassigned OVIRT-1836: ----------------------------------- Assignee: (was: infra) > Optimize options normalization methods in stdci_dsl > --------------------------------------------------- > > Key: OVIRT-1836 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1836 > Project: oVirt - virtualization made easy > Issue Type: Improvement > Components: STDCI DSL > Reporter: Daniel Belenky > > Generally, we have 4 kinds of options: script, globals, options that affect host selection and options to configure the runner. > We need to take care of global options first since they affect everything, then take care of script because it tells which vectors to throw away. > Since runners and pipelines care about different sets of options, we need to split the options handling functions. > This will enable us to compose much more efficient pipelines in the API. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0

[JIRA] (OVIRT-1864) investigate why vdsm check-merged job fails on non el7 hosts
by Barak Korren (oVirt JIRA) 24 Jun '18

24 Jun '18

[ https://ovirt-jira.atlassian.net/browse/OVIRT-1864?page=com.atlassian.jira.… ] Barak Korren reassigned OVIRT-1864: ----------------------------------- Assignee: (was: infra) > investigate why vdsm check-merged job fails on non el7 hosts > ------------------------------------------------------------ > > Key: OVIRT-1864 > URL: https://ovirt-jira.atlassian.net/browse/OVIRT-1864 > Project: oVirt - virtualization made easy > Issue Type: Bug > Components: CI client projects > Reporter: Daniel Belenky > > It seems that vdsm check-merged job fails on non el7 hosts. It just times out. > Maybe el7 hosts are faster(?) > For now, [https://gerrit.ovirt.org/#/c/86780/|https://gerrit.ovirt.org/#/c/86780/] bounds vdsm check-merged to run on el7 hosts only. -- This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087)

1 0