024-09-23 sig-ci weekly meeting - notes
Daniel Hiller
2024-09-23 sig-ci weekly meeting
When: Weekly on Mon, 10:00 – 10:30am
Attendees: Daniel Hiller, Chandra Merla, Brian Carey
Topics:
[urgent]
Daniel Hiller recheck VSOCK test, since it failed last Friday with 4%: https://search.ci.kubevirt.io/?search=VSOCK+communicating+with+VMI+via+VSOCK&maxAge=336h&context=1&type=junit&name=&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job
7% impact, valid for quarantine
revisit previous action items
Daniel Hiller investigate moving cluster-up* into a kubevirtci folder and add an OWNERS file so that we can approve it
started working here: https://github.com/kubevirt/kubevirt/pull/12872
Daniel Hiller run quarantined tests report in the morning - ahead of the meeting: https://github.com/kubevirt/project-infra/pull/3660/files#diff-9a75804be52e179f1dd1d4ff40b225e41280c33c71f84062985677da7444881dR1580
Daniel Hiller quarantined tests report - last storage quarantined tests don’t show the embracing nodes - created issue: https://github.com/kubevirt/project-infra/issues/3664
[every three weeks at beginning of new sprint] do a mini retrospective
propagate the knowledge of fixing flakes
set --output-interceptor-mode=none on the ginkgo call to see output of panicked processes
talk about smells that have been encountered and how to avoid them
how can we prohibit things to enter codebase that we don’t want
discuss imminent topics
[bc] Storage lanes are now running with etcd in memory so we should no longer see etcd timeouts there.
Daniel Hiller some serious clustered failures on the network lanes over the weekend, to add issue and notify network team
Brian Carey to look at the s390x PR from Chandra over the week
Action items
Daniel Hiller q VSOCK test PR: https://github.com/kubevirt/kubevirt/pull/12901
Daniel Hiller to add issue and notify network team - https://github.com/kubevirt/kubevirt/issues/12898
Daniel Hiller to write an email about the k/cluster-up changes , note BREAKING CHANGE
@dhiller to review bc PR
send meeting notes to kubevirt-dev
Kind regards,
Daniel Hiller
He / Him / His
Senior Software Engineer, KubeVirt CI, OpenShift Virtualization
 |
Red Hat GmbH, Registered seat: Werner-von-Siemens-Ring 12, D-85630 Grasbrunn, Germany Commercial register: Amtsgericht München/Munich, HRB 153243, Managing Directors: Ryan Barnhart, Charles Cachera, Michael O'Neill, Amy Ross
Dan Kenigsberg
2024-09-23 sig-ci weekly meeting
When: Weekly on Mon, 10:00 – 10:30am
Attendees: Daniel Hiller, Chandra Merla, Brian Carey
Topics:
[urgent]
Daniel Hiller recheck VSOCK test, since it failed last Friday with 4%: https://search.ci.kubevirt.io/?search=VSOCK+communicating+with+VMI+via+VSOCK&maxAge=336h&context=1&type=junit&name=&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job
7% impact, valid for quarantine
revisit previous action items
Daniel Hiller investigate moving cluster-up* into a kubevirtci folder and add an OWNERS file so that we can approve it
started working here: https://github.com/kubevirt/kubevirt/pull/12872
Daniel Hiller run quarantined tests report in the morning - ahead of the meeting: https://github.com/kubevirt/project-infra/pull/3660/files#diff-9a75804be52e179f1dd1d4ff40b225e41280c33c71f84062985677da7444881dR1580
Daniel Hiller quarantined tests report - last storage quarantined tests don’t show the embracing nodes - created issue: https://github.com/kubevirt/project-infra/issues/3664
[every three weeks at beginning of new sprint] do a mini retrospective
propagate the knowledge of fixing flakes
set --output-interceptor-mode=none on the ginkgo call to see output of panicked processes
talk about smells that have been encountered and how to avoid them
how can we prohibit things to enter codebase that we don’t want
discuss imminent topics
[bc] Storage lanes are now running with etcd in memory so we should no longer see etcd timeouts there.
Daniel Hiller some serious clustered failures on the network lanes over the weekend, to add issue and notify network team
Brian Carey to look at the s390x PR from Chandra over the week
Action items
Daniel Hiller q VSOCK test PR: https://github.com/kubevirt/kubevirt/pull/12901
Daniel Hiller to add issue and notify network team - https://github.com/kubevirt/kubevirt/issues/12898
Daniel Hiller to write an email about the k/cluster-up changes , note BREAKING CHANGE
@dhiller to review bc PR
send meeting notes to kubevirt-dev
--Kind regards,
Daniel Hiller
He / Him / His
Senior Software Engineer, KubeVirt CI, OpenShift Virtualization
--
You received this message because you are subscribed to the Google Groups "kubevirt-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kubevirt-dev...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/kubevirt-dev/CAK%2BeyL4jnkmNOjkqDjcUQEQtr07meUZHa0qffRXsRWGV_dP2ug%40mail.gmail.com.
Daniel Hiller
On Mon, Sep 23, 2024, 15:24 Daniel Hiller <dhi...@redhat.com> wrote:2024-09-23 sig-ci weekly meeting
When: Weekly on Mon, 10:00 – 10:30am
Attendees: Daniel Hiller, Chandra Merla, Brian Carey
Topics:
[urgent]
Daniel Hiller recheck VSOCK test, since it failed last Friday with 4%: https://search.ci.kubevirt.io/?search=VSOCK+communicating+with+VMI+via+VSOCK&maxAge=336h&context=1&type=junit&name=&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job
7% impact, valid for quarantine
revisit previous action items
Daniel Hiller investigate moving cluster-up* into a kubevirtci folder and add an OWNERS file so that we can approve it
started working here: https://github.com/kubevirt/kubevirt/pull/12872
Daniel Hiller run quarantined tests report in the morning - ahead of the meeting: https://github.com/kubevirt/project-infra/pull/3660/files#diff-9a75804be52e179f1dd1d4ff40b225e41280c33c71f84062985677da7444881dR1580
Daniel Hiller quarantined tests report - last storage quarantined tests don’t show the embracing nodes - created issue: https://github.com/kubevirt/project-infra/issues/3664
[every three weeks at beginning of new sprint] do a mini retrospective
propagate the knowledge of fixing flakes
set --output-interceptor-mode=none on the ginkgo call to see output of panicked processes
talk about smells that have been encountered and how to avoid them
how can we prohibit things to enter codebase that we don’t want
discuss imminent topics
[bc] Storage lanes are now running with etcd in memory so we should no longer see etcd timeouts there.
Congratulations!Please share if this makes them fall from their ignominious place at ci-health leaderboard.