Bug 2258071

Hi Lokesh & Mohan,

Im checking in with all F40 change owners now that we are passed the Testable deadline for changes and heading towards beta freeze and the 100% complete deadline on how the change is progressing. Have you any status updates you can share?

Thanks!
Aoife

podman 5 definitely landed (I know, cos it broke ansible-podman and I had to fix it :>) We seem to be at podman-5.0.0~rc1-3.fc40 . So let's call this MODIFIED since we're not at 5.0.0 final yet.

We are targeting a feature-complete rc3 before the deadline. A stable 5.0.0 might not make it before then. Will that pose a problem? 

Matt, Mohan, please correct me if I'm wrong about the dates.

For context, I expect the Linux bits will be finalized and fully stable by RC3, releasing this Thursday, and completely ready to use and test. We're keeping it in RC to continue polishing our Windows/Mac experience.

Just from the mechanistic point of view of the change process I wouldn't call it a problem. FESCo could raise a stink about shipping an RC if they want to, but I doubt they would in the circumstances.

Since the plan is to release with rc3 but we're currently on rc1, MODIFIED seems like the right status. (it's a perfectly cromulent status at this point in the cycle, so don't worry.)

Podman 5 rc2 is part of last night's branched compose (currently composing) as it got pushed to stable yesterday, we are targeting to release rc3 upstream today which should land before the code completion deadline of Feb 27th.

Hi Adam, https://bodhi.fedoraproject.org/updates/FEDORA-2024-1ddc1072a9 hasn't made it past gating being stuck on a nasty kernel bug which has apparently also made its way into f40. You'll see the gating test status says ABSENT. I hear the fix should land in rawhide soon and maybe also f40? I'm hoping an exception will be made for such cases.

Yes, that's fine, you really don't need to worry, this Change is in a perfectly fine state.

Do you have any information on this "nasty kernel bug"? I have been trying to figure out why CI gating tests are missing for some updates all morning.

(In reply to Adam Williamson from comment #8)
> Yes, that's fine, you really don't need to worry, this Change is in a
> perfectly fine state.

That's a relief, thanks :)

 
> Do you have any information on this "nasty kernel bug"? I have been trying
> to figure out why CI gating tests are missing for some updates all morning.

https://lore.kernel.org/all/6a150ddd-3267-4f89-81bd-6807700c57c1@redhat.com/

Oh, wow, I ran into that in a completely different context...https://github.com/containers/bubblewrap/issues/620 . Fix should be in 6.8rc6, which is now stable in f40 and f41, so we just need CI to do whatever is necessary to update and rerun missed tests...

It seems podman 5 has changed the rootless network provider from slirp4netns to passt but this is not mentioned in the change, nor is the migration path documented in the change as a result, are configs automatically migrated from slirp4netns -> passt, are we expected to ship both to enable migration?

There was no migration path planned; existing containers were planned to continue using slirp4netns. Our expectation was that upgrades would continue to include slirp4netns, as it is still a supported network provider (users can manually opt for it with `--net=slirp4netns` and may wish to do so to avoid some quirks of Pasta).

FTR, I see podman-5.0.0-1.fc40.x86_64 in F40 now.

F40 was released on 2024-04-23, so I am closing this tracker. If this Change was not completed, please notify me ASAP.