Date & Time
August, 12th, 2020; 9pm PST
Agenda
Minutes
Release:
- Mg SR2 is fully done and we shipped with a known regression
- Sodium SR4 has release candidate but lots of CSIT failures because of several reasons (aborted infra, python2→3 migration failures in SFC, netconf might have a regression)
- probably looking at next week for SR4 official release
- Luis Gomez and Jamo Luhrsen will continue to look at the CSIT failures
- Aluminium is locked so if we can get all work done and merged we can start thinking about taking a release candidate
- still a lot of open patches though, with most in netvirt
- we need to look at CSIT issues to make sure no more regressions or blockers there
- release checklist is here
- Silicon is now master and we need a schedule to have the TSC approved
- Anil Belur brings up an issue where we lock a branch on code freeze, but some blocker bug that requires an MRI bump then we have
to grant super committer rights which means the branch is unlocked but only with super committer permissions. Now, when super committers
remove their rights again, the branch will stay open unless the release engineer knows that they should lock the branch again.
Infra:
- Jobs are still aborting and Vexxhost thinks it was a problem with the images and we have been updating images, but that does
not seem to be working. The aborts (because instances are not reachable) are not happening on every job- Anil Belur has opened a support ticket with Vexxhost
- Thanh Ha (zxiiro) notices that builder VMs are taking a very long time (10m) to come up
- This issue has been going on for over a week
- A lot of patches were recently merged to update a lot of our infra images
- be wary if jobs are failing in a weird way that maybe some new image issue is there
Docs:
Misc:
Action Items