[01:18] thumper: take another look at https://github.com/juju/juju/pull/10562 plz? [01:45] babbageclunk: ack [01:50] babbageclunk: lgtm with one minor change [01:50] adding a clock to the unit agent manifold config [01:50] there are three other uses of clock.WallClock in that file. [01:54] thumper: ok wilco [02:06] * thumper starts looking at some of these intermittent failures [02:06] * thumper wonders if there is some nifty scripting we could do to scrape jobs to determine frequency of intermittent failures [02:10] sounds like you need a job status prometheus exporter [03:48] thumper: I keep meaning to write something to do that but getting distracted [03:49] babbageclunk: yeah, understood [03:49] would be super useful though [03:54] hpidcock: when thumper is done with you.... https://github.com/juju/juju/pull/10563 [04:23] crap... [04:23] my fix didn't [04:40] * thumper pokes some more [04:50] wallyworld: looks good, few comments that's all [04:50] ty [04:54] hpidcock: because the v3 CLI is strictly opt in, there's no need to be too verbose about reminding users what the v2 syntax is [04:56] wallyworld: yeah, I understand that, but I imagine at some point v3 won't be opt in. Was just trying to highlight that we should provide feedback when something is deprecated. [04:58] i can add a note to the help text for "juju run" [05:00] I think I have it now [05:00] over 200 concurrent successes without a failure and during a stress run of the entire test suite [05:00] * thumper pushes [05:03] wallyworld: https://github.com/juju/juju/pull/10564 [05:03] ok [05:15] thumper: lgtm [05:15] wallyworld: ta [05:44] landing that branch, and EOD [05:44] later peeps [09:35] jam, got a sec? [09:40] jam, scratch that, i've worked it out [10:11] can someone do a quick CR on https://github.com/juju/charm/pull/290? [10:12] jam: ^^ is the fix for #1841105 [10:12] Bug #1841105: panic: runtime error: invalid memory address or nil pointer dereference [11:00] achilleasa: Looking. [11:03] achilleasa: approved. [11:18] manadart: can you also please check https://github.com/juju/juju/pull/10565 (brings the charm.v6 changes to juju) and try the QA step? [11:28] achilleasa: Sure, [12:35] manadart: review please: https://github.com/juju/juju/pull/10555 [12:44] hml: Looking. [12:51] guild in 9 [14:21] hml: What provider are you using to add subnets? [14:22] manadart: :-) I used maas, but the call failed because the subnets were already created. so i made sure that the correct facade versions were being used and the errors stayed the same. i’ve since learned that aws might work there better. [14:29] hml: OK, yeah. I was using AWS. [15:12] manadart: you had no issues bootstrapping a aws on a develop branch today? [15:14] rick_h, omitempty won't work for us, as the bundlechange api doesn't work the way other apis work [15:41] stickupkid: doh, otp atm but let's table it then. We don't need to solve it until 2.7 [15:41] stickupkid: and go with the release for the bug affecting the customer please [15:41] rick_h, ok sounds good [15:41] stickupkid: and we can look into what it'll take to have bundlechanges behave like a real boy api [16:01] rick_h, https://github.com/juju/python-libjuju/pull/349 release bump [16:03] stickupkid: k, +1'd [16:03] rick_h_, ty [16:21] hml: If you are interested, I have started to put together https://github.com/juju/juju/pull/10566. It will accrue context as I push the other pieces up. [16:21] manadart: cool - will start to look === hml_ is now known as hml [20:54] morning