[01:05] axw: quick HO? 1:1 [01:07] thumper: you around? [01:10] wallyworld: just got back, brt [01:19] axw: take a look at https://github.com/juju/juju/pull/7547? [01:19] wallyworld: too late? [01:19] babbageclunk: no, there's a unit test failure we're trying to fix [01:19] i'll look [01:20] wallyworld: thanks [01:24] babbageclunk: yeah, i think that looks good. a small surface area [01:27] wallyworld: cool, thanks - hopefully makes it in! [01:27] it will :-) [01:29] axw: interestingly, i now see the test fails without --race but succeeds with --race. it seems likely related to the mutex i added to fix the race failures [01:30] wallyworld: it doesn't matter whether you use -race or not, it's just time sensitive. I *think* the issue is due to flushing on ErrDying [01:30] not sure tho yet [01:30] ok, i'll check back in soon hopefully [01:31] axw: you may need to pull my branch and re-propose if you get a fix done before i get back [01:31] wallyworld: ok [01:31] tyvm [01:33] axw: yeah, if we don't flush on dying that test passes and another fails [01:34] wallyworld: I think the test is kinda bogus, and shouldn't assume it'll return an error [01:35] and now i can't get it to faio anymore [01:36] *fail [01:36] axw: i kinda agree [01:37] i might just tweak the test not to check for that error [01:37] it still checks that sync returns on shutdown [01:39] axw: done that, will land now [01:39] wallyworld: we should probably have another tests that checks that that error is returned after the PingBatcher has completely stopped [01:40] probably [01:40] done [01:42] wallyworld: for now [01:42] thumper: tis ok [01:42] ack [01:43] axw: landing now, could you keep an eye out while i pop out and remerge if there's a spurious failure? [01:44] babbageclunk: your pr failed [01:45] wallyworld: I saw - looks spurious [01:45] faaaark [01:45] yeah it does :-( [01:46] wallyworld: yep np [01:46] Running it under race detector now just to see [01:46] babbageclunk: tempted to merge by hand [01:46] so we can unblock the release [01:46] but i gotta run out for an hour] [01:47] veebers: burton-aus: race fix landing now. there's another pr from babbageclunk which just failed with a spurious error - we could merge that directly since CI will run the tests anyway [01:48] i'll check back in a bit [01:50] wallyworld: ack sounds good. Let us know which lands last, we'll check the build make sure everything is in there and watch the CI run [02:02] babbageclunk: I have to go out for a little while too, can you please let veebers know when your branch has landed? wallyworld's just finished [02:13] axw: wilco [02:16] Everyone's going out. I think I'll go out in a bit. [02:19] anyone able to give me a quick update on https://bugs.launchpad.net/juju/+bug/1676279 ? [02:19] Bug #1676279: canceling destroy-model prevents further calls to destroy-model to succeed [02:19] anastasiamac: ^ [02:22] * anastasiamac looking [02:22] blahdeblah: yes, it would have been included in 2.2.0 as it was committed back in time for 2.2-b3.. I'll answer on the bug as well [02:24] thanks anastasiamac [02:24] also, any idea when 2.2.1 will be out? [02:25] Also, any clarification on https://bugs.launchpad.net/juju/+bug/1699050 as to whether the workaround which wgrant tested needs to be applied to each model? [02:25] Bug #1699050: remove-application regression on 2.1 -> 2.2 upgrade with subordinates [02:25] babbageclunk: ^ [02:33] veebers: Hey, my PR's just landed: https://github.com/juju/juju/pull/7547 [02:33] blahdeblah: aiming to release 2.2.1 today (last i've heard) [02:36] babbageclunk: that's the one for the last bug above? [02:36] blahdeblah: yup [02:36] cool - thanks [02:36] Anyone able to comment on https://bugs.launchpad.net/juju/+bug/1669046 as well? [02:36] Bug #1669046: juju-db very high load on primary controller unit [02:37] babbageclunk: sweet [02:37] blahdeblah: I was answering your second question. (Just in case you weren't sure.) It'll be in 2.2.1. [02:38] yep [02:38] babbageclunk: that went in after Ians pingbatcher fix right? [02:38] blahdeblah: so hopefully you don't need to apply the workaround? [02:38] veebers: I think so - just checking... [02:38] babbageclunk: Once 2.2.1 comes out? [02:38] veebers: yup [02:38] blahdeblah: yes [02:39] We're hoping that 2.2.x will resolve a number of our load issues, but since it was never confirmed that the root cause in 1669046 is 1676279, I wonder whether the former needs more diagnosis. [02:42] babbageclunk: cool, once it gets a revision build we'll monitor it for release [02:49] Are we going to do a mass-merge from 2.2 to develop at some point, or should I be forward-porting things myself? [03:14] * babbageclunk goes for a run for a bit [03:27] babbageclunk: no, we will merge 2.2 into develop. i was waiting for all the things to land [04:33] wallyworld: yeah, I thought we'd probably be doing that again [05:06] babbageclunk: yeah, all done :-) [05:07] wallyworld: nice one! [05:32] Booking my shuttle for the sprint - when are people mostly leaving the place? Don't want to be hanging out all by myself in Mooloolaba when everyone's already left! My flight's not until 7. [05:32] (on the Saturday) [05:33] babbageclunk: I've only booked my shuttle from airport -> venue. Was going to book the return during the week [05:34] veebers: but you're missing out on a $6 saving! [05:35] babbageclunk: hah true. But something better might pop up. In the past I've been able to hitch rides with people heading to the airport etc. Although I geuess this time around it's a bit more of a trip [05:36] veebers: Actually, I like your plan, I'm going to do the same. [05:36] And then say it was my idea all along. [05:40] babbageclunk: well, if we get stranded and miss our flights you can take all the credit you want ^_^ [05:40] Yay! [05:41] babbageclunk: any idea how many hours (or days) your supposed to be at BNE before your flight? [05:41] Oh no - that's a good point, I'd forgotten to factor that into my return booking anyway. [05:42] hmm [05:43] * veebers needs to make sure he has enough clean cloths to pack [05:43] Will have to break the shorts out of storage, that is if it's not frozen shut [05:48] You folks aren't coming in via MCY? [05:48] Much nicer :-) [05:49] blahdeblah: I'm just glad there is a direct flight from Dunners and I don't need to be waiting around for hours between flights :-) [05:49] Yeah - pretty unlikely to get MCY direct to there, I guess. [05:49] veebers: BTW, standard in .au is 2 hrs for intl. flights; they might be a bit more lenient with NZ flights - not sure. [05:50] (that was re: being there before your flight) [05:50] blahdeblah: ack thanks. That's about what I expected [05:50] (considering the flights only 4 hours :-P) [06:02] wallyworld: whenever you're free, https://github.com/juju/juju/pull/7550 and https://github.com/juju/juju/pull/7551 [06:02] ok [06:15] wallyworld: so how do you did the forward port? branch off develop, merge 2.2, revert whatever things you don't want like version bump? [06:16] axw: yeah. i only normally revert the version number in 3 places [06:19] babbageclunk: i think we can close pr 7542 [06:19] as it is obsolete with the new, smaller fix [08:42] wallyworld: Yup yup === frankban is now known as frankban|afk