[CI:IMG] Bump crun to 1.14 on CI VMs #6869
Conversation
|
@edsantiago yes thank you, I've got a PR kicking around to address this. |
|
xref #6822 |
edsantiago
force-pushed
the
ci_update_crun_to_1_14
branch
from
7d1f624 to
d5e40ff
Compare
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: edsantiago The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
edsantiago
force-pushed
the
ci_update_crun_to_1_14
branch
2 times, most recently
from
a57236e to
2961aaf
Compare
| @@ -59,7 +59,7 @@ case "${OS_RELEASE_ID}" in | |||
| workaround_bfq_bug | |||
|
|
|||
| # HACK: Need Conmon 2.0.17, currently in updates-testing on F31. | |||
| dnf update -y --enablerepo=updates-testing conmon | |||
| $BIGTO dnf update -y --enablerepo=updates-testing conmon | |||
There was a problem hiding this comment.
This script is more intended for small adjustments that need to happen at runtime. Since you're building new VM images here, might as well include this change into: contrib/cirrus/packer/*_packaging.sh. Currently fedora is setup (I believe in contrib/cirrus/packer/fedora_setup.sh) to NOT use updates-testing on F31. I'm not opposed to undoing that logic so updates-testing is used everywhere.
FWIW: I'm also not opposed to having this live in setup_environment.sh for the short-term. Whatever you think is easiest/best.
There was a problem hiding this comment.
@cevich when you say "might as well include this change", what are you referring to? Do you mean my change, that is, the addition of $BIGTO to a flaky command? Or do you mean e0d9404, the dnf command itself added 2020-06-02 to get conmon 2.0.17? The reason behind my change (adding $BIGTO) is that CI jobs have been failing frequently in this particular step. I wanted a quick band-aid for that, without having to look into all VMs and distros to see if conmon 2.0.17 has made it into stable everywhere.
Since there seem to be multiple CI:IMG PRs right now, and since CI is hosed anyway, I'm going to let this PR sit idle for now; I will revisit once things stabilize.
There was a problem hiding this comment.
I wanted a quick band-aid for that,
Ahh okay, then setup_environment.sh is fine, and you probably don't even need to go through the CI:IMG shenanigans either. My other comments are !quick-band-aid, so it's okay to ignore them 😄
This began as an empty commit intended solely to get CI to rebuild the VMs using crun 1.14 (up from 1.13). Twin goals: 1) Be able to test containers#6693 (--sdnotify option); and 2) Get rid of 'cgroup.freeze' CI flakes. These are fixed by crun PRs 419 and 423 respectively. CI failed on the original PR submission, with errors on ginkgo install. The change to Makefile is intended to address that. The change to setup_environment.sh is intended to address a flake we're frequently seeing with the Fedora dnf repos. This adds a retry in case of a failing dnf command. Signed-off-by: Ed Santiago <[email protected]>
edsantiago
force-pushed
the
ci_update_crun_to_1_14
branch
from
2961aaf to
d23fd91
Compare
|
Whew... in addition to the ginkgo mess, CI is failing because there's a test for a item_test "The varlink executable is present" -x "$(type -P varlink)" || let "NFAILS+=1"Closing again; this is beyond my abilities. |
github-actions
bot
added
the
locked - please file new issue/PR
This is an empty commit intended solely to get CI to rebuild
the VMs using crun 1.14 (up from 1.13). Twin goals:
These are fixed by crun PRs 419 and 423 respectively.
Signed-off-by: Ed Santiago [email protected]