2026-02-25 sig-ci quarantine catch-up - notes

[Daniel Hiller]

2026-02-25 sig-ci quarantine catch-up

When: Weekly on Wed, 9:45 – 10:15am

Notes: KubeVirt CI SIG meeting notes

Attendees: dhiller, dollierp, Dylan White ycui, Federico Fossemo

Reminders:

• we will create GitHub issues for tracking

• GitHub issues and PRs

• should be marked with /sig ci and /kind flake if applicable

• should be marked with the target sig

• kubevirt quarantine process documentation

Topics:

• [urgent]

• 
  

• ci-health

• 

• 

• previous action items

• state of existing issues: https://github.com/kubevirt/kubevirt/issues?q=is%3Aissue+is%3Aopen+label%3Akind%2Fflake+sort%3Aupdated-asc+label%3Asig%2Fci

• [non-urgent]

• [dhiller] it was reported that a periodic was suffering from soft lockups (source)

• we noticed that the periodics aren’t spread efficiently - a couple of those were started in the same time range, which likely caused cpu saturation on the note where the affected periodic was running

• 

• PR https://github.com/kubevirt/project-infra/pull/4742

• PR https://github.com/kubevirt/project-infra/pull/4744

• 
  

• Look at flakes

• flake stats - create issues accordingly

• overall ( ∑=3350, 100.00% )

• pull-kubevirt-e2e-k8s-1.35-sig-compute-serial ( ∑=755, 22.54% )

• periodic-kubevirt-e2e-k8s-1.35-sig-compute ( ∑=613, 18.30% )

• pull-kubevirt-e2e-k8s-1.34-sig-compute-serial ( ∑=541, 16.15% )

• periodic-kubevirt-e2e-k8s-1.33-sig-compute ( ∑=536, 16.00% )

• periodic-kubevirt-e2e-k8s-1.34-sig-compute ( ∑=272, 8.12% )

• periodic-kubevirt-e2e-k8s-1.32-sig-compute ( ∑=114, 3.40% )

• periodic-kubevirt-e2e-k8s-1.34-sig-performance-kwok-100 ( ∑=65, 1.94% )

• periodic-kubevirt-e2e-k8s-1.35-sig-storage ( ∑=55, 1.64% )

• periodic-kubevirt-e2e-k8s-1.34-sig-monitoring ( ∑=43, 1.28% )

• periodic-kubevirt-e2e-k8s-1.34-sig-operator ( ∑=42, 1.25% )

• pull-kubevirt-e2e-k8s-1.35-sig-operator ( ∑=40, 1.19% )

• (new) VSOCK test failures https://storage.googleapis.com/kubevirt-prow/reports/flakefinder/kubevirt/kubevirt/flake-stats-14days-2026-02-25.html#%5bsig-compute%5dVSOCK%20should%20return%20err%20if%20no%20app%20listerns%20on%20the%20port

• fossedihelm to take a look

• Result: It’s part of the clustered failures

• (new) https://github.com/kubevirt/kubevirt/issues/16936

• dequarantine tests:

• look at list of quarantined tests

• count: tests in quarantine currently
  

• check status, i.e. who is working on those

• look at PRs that want to fix flakes

• see whether we can dequarantine tests 

• misc topics

• [fossedihelm] per next release

• before code freeze we need to prioritize merging PRs against main vs. merges against release branches

• use approved-vep label as criteria for prioritizing the merge

• after code freeze and creation of latest release branch we need to prioritize merging against latest release branch

• TODO find issue that describes in more detail

• look at what k8s does

Action items

• 
  
• 

  update/create issues with latest flakes spotted

• 

  communication

• send meeting notes to kubevirt-dev, bcc sig people for spotted flakes (include meeting changes for upcoming instances)

--

Kind regards,

Daniel Hiller

He / Him / His

Principal Software Engineer, KubeVirt CI, OpenShift Virtualization

Red Hat

dhi...@redhat.com   

Red Hat GmbH, Registered seat: Werner von Siemens Ring 12, D-85630 Grasbrunn, Germany  
Commercial register: Amtsgericht Muenchen/Munich, HRB 153243,
Managing Directors: Ryan Barnhart, Charles Cachera, Avril Crosse O'Flaherty