Upstream has responded that they have seen the pull request, and they are going to assign it to one of their engineers to review:
> Thanks for reporting and putting together a patch - the engineering team in > EC2 that maintains the agent is aware of the patch and are prioritizing > testing it. I'll ask them to give you an ETA when they have one. https://github.com/aws/amazon-ec2-hibinit-agent/issues/20#issuecomment-1104679221 Do you want to wait until we get a positive review from upstream? I suppose we can tell AWS that the case can be WoCus while we are waiting for upstream to review, and it will pause the SLA timer. On Thu, Apr 21, 2022 at 5:54 PM Matthew Ruffell <matthew.ruff...@canonical.com> wrote: > > Hi everyone, > > Could you please review LP #1968805 [1], and sponsor the uploads if it looks > okay? > > [1] https://bugs.launchpad.net/ubuntu/+source/ec2-hibinit-agent/+bug/1968805 > > I think bumping the priority to 32767 is the best solution for this particular > case, since it would be higher than priorities users would be using in the > wild, > if they happen to have multiple swapfiles configured. > > I have submitted the patches upstream, but no word back. The upstream is AWS > though, so maybe I might be able to put the SLA on hold for feedback there, > but otherwise, we have about one month left on the SLA to get this fixed, so > I wasn't going to wait. > > I did ask the CPC team about their feelings on the patches, but I didn't get > much of a response other than from Chris Newcomer in the CPC channel. > > If you were going to test this for yourself, best stick with Focal for the > moment, as Jammy is broken on xen instance types, and is being tracked > separately in https://bugs.launchpad.net/ubuntu/+source/linux-aws/+bug/1968062 > > Let me know if you have any feedback, or think this should be fixed in a > different way. > > And yes, Dan, this is probably a duplicate of > https://bugs.launchpad.net/bugs/1910252 > where the real root cause is that systemd blindly hibernates to the highest > priority swapfile, as it has no way to know what resume device and offset the > kernel is configured to resume from. SRUing such a change might be difficult > as those who accept the standard behaviour would have to manually update their > configuration on their systems to tell systemd what swapfile to hibernate to. > > I think talking with upstream systemd will go past the 1 month left on the > SLA, > and so these straightforward patches to ec2-hibinit-agent are probably the > best > low risk way to work around the bug, and fix AWS users. > > PS, our ec2-hibinit-agent package diverged from upstream long ago, and the > foundations team appear to be happy carrying patches not upstreamed. > > Thanks, > Matthew -- Mailing list: https://launchpad.net/~sts-sponsors Post to : sts-sponsors@lists.launchpad.net Unsubscribe : https://launchpad.net/~sts-sponsors More help : https://help.launchpad.net/ListHelp
