[00:16] axw: can you take a gander at https://github.com/juju/juju/pull/6868? [00:37] babbageclunk: looking [00:43] babbageclunk: all the other works are quiesced while this code runs, right? [00:43] workers* [00:44] babbageclunk: the provisioner won't be trying to update things, for example? [00:49] menn0: axw: for bug 1660087, we just need to know if it has been addressed in later revisions, on 2.1.x [00:49] Bug #1660087: Controller and all models unresponsive [00:50] anastasiamac: can't say until we know what the root cause is [00:50] axw: hopefully, logs attached to the bug will help :D [00:51] anastasiamac: yeah, hopefully. I'll look when I can [00:51] axw: nps \o/ [00:57] axw: yeah, that's right - they're all gated by the migration flag [00:58] babbageclunk: ok, thanks. I've left some comments [00:58] axw: cool, thanks! [01:04] anastasiamac: sorry I was out for a bit [01:04] anastasiamac: what axw said :) [01:04] menn0: thnx :) [01:19] * menn0 is going dark for a while to focus on CAAS [01:32] * redir goes eod [01:34] * redir will look at #1654603 in the morning if still waiting for PPC64EL MAAS access. [01:34] Bug #1654603: kill-controller takes too long, is it waiting for hooks to finish? [02:25] Does anyone know how to get maas to see a change in max ram after changing it in vmm? [02:26] Do I need to remove and recommission it, or is there a more direct way? [02:32] babbageclunk: i wonder if maybe we should reach out to MAAS guys? dunno if any of them are in this channel.. [02:33] anastasiamac: yeah, I'll go ask over there. [04:47] Bug #1660522 opened: Un-removeable "life: dead", "agent-state: pending" agents

[05:24] jam: bridge testing again, didn't notice IRC failed to reconnect. wallyworld raised the same issue of remove-storage vs. detach-storage [05:25] jam: my question is what that means in the case of shared storage [05:26] jam: we can just have it mean "detach from the application", which might be reasonable? [05:51] axw: sorry, I've got a bunch of different things in my head right now. I just wanted to make sure I understood what made you uneasy, so that I understand the full problem. I probably won't be able to switch to that for a bit [05:52] jam: it's all good, I'm looking at the lxd/maas networking issue now [07:01] wallyworld: ping? [07:01] hey [07:01] wallyworld: yo. about the firewaller thing. can you HO? might be easier [07:02] sure, give me 5 [07:02] wallyworld: np, ping when [07:09] 1:1? [07:20] wallyworld: yes [07:35] Bug #1660542 opened: container mac addresses should use 'locally assigned' section

[07:35] Bug #1660543 opened: container mac addresses should use 'locally assigned' section

[07:41] axw: changes pushed [07:41] wallyworld: looking [07:44] wallyworld: LGTM [07:44] ty === frankban|afk is now known as frankban [08:26] axw: take another look? https://github.com/juju/juju/pull/6868 [08:26] babbageclunk: looking [08:27] axw: thanks! you weren't kidding when you said callAPI was ugly. [08:27] babbageclunk: yeah :/ [08:27] babbageclunk: probably should just bite the bullet and write our own wrapper API clients [08:28] (not in this PR though) [08:28] axw: no :) [08:30] babbageclunk: done [08:30] axw: cheers === wesleyma` is now known as wesleymason [11:00] Bug #1660543 changed: container mac addresses should use 'locally assigned' section

[12:30] morning all [13:48] redir: ping when/if you're about [13:55] rick_h: I doubt he will be here for a couple of hours [13:55] perrito666: yea, just realized. I ping'd him out of band [13:55] * rick_h somehow thought the day was farther along [14:15] a review of https://github.com/juju/juju/pull/6893 is requested [14:15] it involves refactoring the code that checks what bridges to create for containers, but doesn't, itself, change any of the logic [14:15] just pulls it off of Machine where it didn't really belong. [14:15] frobware: ^^ [14:59] jam: reviewed [15:07] frobware: responded and pushed up an update [15:07] care to take another look [15:08] jam: looking [15:09] jam: LGTM [15:14] jam: I was just taking another look on the files that were added. [15:17] jam: those tests - they were just the tests that were moved? [15:19] jam: if the error doesn't bubble up to the user then I think what you suggested is fine. [15:20] frobware: so how the code is *currently* used, the only caller of it ensures that its only dealing with bridges first [15:20] frobware: I added the check because I split the code from where it was created from where the check was done [15:20] frobware: and I believe in validating inputs in case someone uses it differently in the future [15:21] frobware: hypothetically, that sort of error could bubble to the user if we have a bug in our software, so I did tweak the wording [15:21] jam: ack [16:51] rick_h: pong [16:51] I see a task in the Juju Show [16:52] redir: howdy, yep [16:52] todo made [16:52] redir: wanted to see if you had some details/etc you wanted/could flesh out [16:52] redir: ty [16:52] rick_h: not much, uvtool is gone and it works on ARM64 where kvm does on ARM64 now. [16:53] redir: and works everywhere else expected and this is 2.2 ? [16:53] otherwise there's no change to the API or feature set, but I can make an outline [16:53] rick_h: still trying to get access to ppc64el HW to test and debug there. === mskalka is now known as mskalka|lunch [17:39] brb need to let the tree trimmers in the garden === mskalka|lunch is now known as mskalka [18:15] brb, afternoon snack :) === frankban is now known as frankban|afk [19:42] perrito666: you made me hungry [19:42] redir: sad to be you mate :p [19:42] I actually had to drive 15 min to get the snack [20:22] develop is broken [20:34] same breakage in 2.1 [20:46] redir: what's the breakage? (I haven't looked) [20:47] # github.com/juju/juju/worker/machineactions [20:47] worker/machineactions/handleactions.go:70: unknown "github.com/juju/utils/exec".RunParams field 'User' in struct literal [20:47] worker/machineactions/handleactions.go:87: not enough arguments to return [20:47] menn0: ^ was going to look at the history after lunch [20:47] how the hell did that get in? [20:48] the commit where that went in was the 26th [20:48] perhaps before verify.bash was enabled on CI [20:49] s/enabled/fully enabled [20:49] and someone prolly didn't have the pre-commit hook enabled [21:03] redir: seems likely [21:06] OK nm, I think It is me [21:06] shockingly [21:08] I fatfingered godeps update and missed the error === JoseeAntonioR is now known as jose === mskalka is now known as mskalka|afk [22:10] menn0: as an OCR, this PR addresses frequent intermittent bug: https://github.com/juju/juju/pull/6892 [22:33] jam: YUNO squash? [22:48] perrito666: prolly because there were conflicts and or merges [22:55] juju deploy charm1 charm2 # is this valid syntax? [23:16] it appears not but it deploys charm2 [23:28] but claims it deployed charm1 [23:44] be a couple minutes late to s/u [23:46] redir: that sounds like sensible behaviour :-/