Wednesday, 2021-06-30

kata-irc-bot<liubin0329> @bergwolf @fidencio is it possible make it as non-required temporarily only for this PR and let it get to be merged ?09:32
kata-irc-bot<fidencio> I even suggested that in the PR09:33
kata-irc-bot<fidencio> but I didn't get any answer from the other members of the AC09:33
kata-irc-bot<bergwolf> sounds good to me @fidencio09:40
kata-irc-bot<samuel.ortiz> SGTM as well @fidencio09:42
kata-irc-bot<fidencio> Done!09:49
fidenciodevimc: oooi! so, you think all the instability in the CI is because of the switch to q35?13:22
devimcfidencio, I think so13:22
devimcI'll give a try13:22
fidenciodevimc: I can see the metrics being affected, but more than that (as in the stability tests breaking and whatnot) seems suspicious13:23
fidenciodevimc: do you need some help from marcel on this one?13:23
devimcfidencio, not yet13:23
devimcbut thanks13:24
fidenciodevimc: sure, please, feel free to poke us for anything needed13:24
devimcsure thing :)13:24
kata-irc-bot<fidencio> @marcel, JFYI ... ^^^13:24
kata-irc-bot<christophe> It's weird that the PR CI did not catch this. Also, is there a way to recover the CI testing for a merged PR in GitHub, or are the results lost forever?14:31
kata-irc-bot<fidencio> It's weird and it's already raised https://github.com/kata-containers/kata-containers/pull/2127#pullrequestreview-693736635 ... @wmoschet even did some preliminary investigation on that.14:53
kata-irc-bot<fidencio> About recovering the CI after a PR gets merged ...14:54
kata-irc-bot<fidencio> Simply click on details, all the info is there.14:54
kata-irc-bot<fidencio> The info will disappear after a month or so, but that's usually enough time.14:54
kata-irc-bot<marcel> @fidencio @julio.montes please let me know how I can help15:23
kata-irc-bot<fidencio> Not much to do at this point, reverting the changes didn't help with the issues we're seeing, so @julio.montes just closed the PR.15:24
kata-irc-bot<fidencio> But if you happen to have any cycles to help us to try to figure out why the CI went awry since a few days ago, that would be very much appreciated.15:24
kata-irc-bot<marcel> great, I ll19017941263915:24
kata-irc-bot<marcel> great that it was not that PR fault!15:25
kata-irc-bot<marcel> @fidencio I can try to help, sure, please point me to the first place to look ... :slightly_smiling_face:15:25
kata-irc-bot<marcel> I suppose I can choose a PR and look for CI logs, right?15:26
kata-irc-bot<fidencio> This is one of the issues, and there's some discussion there.  @wmoschet is atop of some the issues, but there are others failing (as he and @bergwolf pointed somewhere in that discussion)15:27
kata-irc-bot<marcel> @wmoschet @bergwolf please point me to some specific issue and I'll start digging15:33
kata-irc-bot<bergwolf> https://github.com/kata-containers/tests/issues/3654#issuecomment-87128255615:33
kata-irc-bot<marcel> @bergwolf thanks! Let's see if I can find anything15:41
kata-irc-bot<marcel> Well... I am kind of lost :( .ll19017918558016:34
kata-irc-bot<marcel> Is it possible to run the CI, or at least a part of the CI in a VM on my machine? I would rather reproduce then guess based on the logs...16:37
kata-irc-bot<fidencio> There's, you can check the parameters passed to the jenkins job build script (which is part of the tests repo), and from that run it locally16:42
kata-irc-bot<fidencio> I'd recommend a new VM ...16:42
kata-irc-bot<fidencio> @marcel, but that's for tomorrow. :slightly_smiling_face:16:42
kata-irc-bot<marcel> @fidencio if you''l have the time to "show" me tomorrow, I'll be happy to achieve that skill ...16:43
kata-irc-bot<marcel> or any other day :slightly_smiling_face:16:44
kata-irc-bot<fidencio> Sure, I have a doctor appointment around 9am (your time) but I'll ping you once I'm back at home. :slightly_smiling_face:16:44
kata-irc-bot<marcel> thanks16:44
kata-irc-bot<fidencio> thanks!16:44
kata-irc-bot<wmoschet> @marcel regarding reproduce CI job locally in a VM, see this https://github.com/kata-containers/tests/pull/3547  Please, post your feedback (if any) there17:30
fidenciojcvenegas++18:56
fidenciojcvenegas++ (as the bot is back again)18:56

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!