On 1/12/23 21:09, Dumitru Ceara wrote: > On 1/12/23 21:08, Mark Michelson wrote: >> On 1/12/23 15:03, Dumitru Ceara wrote: >>> On 1/12/23 20:58, Mark Michelson wrote: >>>> I made the change Dumitru suggested (checkout@v3 instead of checkout@v2) >>>> and merged both patches to main. >>>> >>>> Thanks Dumitru and Ales! >>>> >>> >>> Thanks, Mark, but the second patch should not be pushed to branches >>> newer than 22.09. We should revert it on main because we actually want >>> to run the ovnkube session affinity tests there and skip them on >>> branches <=22.09. >>> >>> I can take care of the backports tomorrow, but it would be great if you >>> could already revert the patch on main branch. >>> >> >> Sorry about that. I pushed a commit to revert patch 2 on main: >> https://github.com/ovn-org/ovn/commit/3abc62da8d323c52cfbf7f7dbad9cd50eb9afdcc >> > > Thanks for the quick response!
I pushed the corresponding patches to all stable branches down to 22.03. Unfortunately upstream ovn-kubernetes CI is flaky again so we still have some control-plane test failures: https://github.com/dceara/ovn/actions/runs/3910245152/jobs/6682549215#step:12:12531 Tim, Surya, do you maybe have some time to help us determine why these fail so often? I don't think anything major changed in OVN lately that would cause these to fail. The session affinity timeout conformance tests OTOH are skipped/run as needed: https://github.com/dceara/ovn/actions/runs/3910245152/jobs/6682548617 (skipped) https://github.com/dceara/ovn/actions/runs/3910234735/jobs/6682600435 (executed) Best regards, Dumitru _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
