samueldr changed the topic of #nixos-infra to: NixOS infrastructure | logs:
nh2[m] has joined #nixos-infra
<nh2[m]> I'm a bit at a loss btw where we track infra issues, sometimes there are notifications on top of, sometimes Hydra issues are pinned in the nixpkgs issue tracker, but I feel like it's pretty ad-hoc.
<nh2[m]> It would be nicer if there was a dedicated issue tracker for this, vs reading logs of multiple IRC channels (unless there already is one of course).
<nh2[m]> For example, 20.09 doesn't seem to have Thunderbird or Libreoffice currently, and for Thunderbird it's because of some SSH error to a builder machine:
cole-h has joined #nixos-infra
<samueldr> at first glance it looks like the job just missed being built
<samueldr> not sure if there is a larger issue here
<samueldr> restarted the build
<samueldr> (failure state can still be seen here )
<samueldr> ah, blind me, it timed out
<samueldr> I didn't look past the first error listed
<samueldr> or uh, past that abort*
<samueldr> I wouldn't expect a tracking issue for a simple time out, as annoying as it can be
<samueldr> yeah
<samueldr> I think users sometimes do report it to the nixpkgs tracker when a package has unexpectedly failed on Hydra
<samueldr> I don't have an answer
<nh2[m]> samueldr: IMO unless something speaks _against_ it, we should just use the nixpkgs repo with an `build-infa` label, and write on the top of and something like "If a package is not building due to what looks like an infrastructure issue, check or report it [here](" and extend the Github issue
<nh2[m]> templates by one for "Stuff isn't building on Hydra"
<samueldr> I don't have a strong opinion about that
<nh2[m]> I'll wait a while and see if anybody pops up saying that this would be bad, if nobody does I may go make some PRs for that
<gchristensen> if we do that, maybe should extend the page to show those in addition to issues labeled channel blocker
danderson has joined #nixos-infra
jonringer has joined #nixos-infra
<cole-h> We do have a general purpose "infrastructure" label that could be repurposed?
<nh2[m]> cole-h: that sounds good!
<nh2[m]> gchristensen: yeah that's a good idea
LnL- has joined #nixos-infra
LnL- has joined #nixos-infra
LnL- has quit [Changing host]
LnL has quit [Ping timeout: 260 seconds]
cole-h has quit [Ping timeout: 264 seconds]
henrietaMag has quit [Quit: ZNC 1.7.5 -]
lassulus has joined #nixos-infra
henrietaMag has joined #nixos-infra
jonringer has quit [Ping timeout: 264 seconds]
aristid has joined #nixos-infra
<MichaelRaskin> Should the label be (eventually) auto-added on creation if Hydra issue template is used (with triager maybe removing it later)?
jonringer has joined #nixos-infra
cole-h has joined #nixos-infra
<nh2[m]> MichaelRaskin: I think so
cole-h has quit [Ping timeout: 264 seconds]
<jonringer> the tests fail in like.. 2s, but they don't clean up files correctly, so it's just in an invalid state until the job times out
<worldofpeace> jonringer: u should have permissions to do this
<worldofpeace> I cancelled them
<jonringer> thanks
<jonringer> PR to fix the underlying issue
MichaelRaskin has quit [Quit: MichaelRaskin]
<worldofpeace> jonringer: lgtm
<jonringer> Oh, my email does have permissions
<jonringer> well, time to restart a few hundred tests
<worldofpeace> jonringer: IIRC amine turned on configuration edits as well
<gchristensen> you could restart all failed if you wanted
<jonringer> oh, i just wanted to do aborted, but that's another option I'm able to do. :)