[04:32] If anyone fancies doing a review, take a look at https://github.com/juju/juju/pull/6622, thanks! === axw_ is now known as axw [09:45] mornin' all === frankban|afk is now known as frankban [10:05] mgz: G'day, is there a magic touch for PR test retries that you have? [10:05] https://github.com/juju/juju/pull/6620 [10:17] * frobware needs to step out for ~1 hour. [10:32] Morning, is there a configuration on 2.x where you could configure an openstack cloud deployment to use by default floating IPs on the controller? Something like https://jujucharms.com/docs/1.25/config-openstack ? I can't seem to be able to find any reference for 2.0 ? [11:24] deanman: I think you should be able to pass that to bootstrap as a --config option. [11:25] macgreagoir: yeah i just found out by peeking the Go code, there is a use-floating-ip configuration option which i wasn't sure whether is implemented in 2.0. I use https://jujucharms.com/docs/stable/models-config as a reference for available keys. [11:26] deanman: I was just peeking at the provider code too :-) [11:29] macgreagoir: (y) [12:17] macgreagoir: so, cleanup done, though you upset the checkjob by targetting against staging first [12:18] I have poked, and have another thing to try as well if needed [12:18] mgz: Guilty! [12:30] morning [14:47] fnordahl anastasiamac mgz: juju/juju/pull/6617 merged successfully. Thanks fnordahl! [14:49] macgreagoir: thank you! [15:13] rick_h: what was the windows thing you wanted me to look at? [15:14] natefinch: should have a GH email from the PR [15:20] rick_h: ahh, ok, I forgot I finally fixed it so github notifications for juju go to my canonical address and not my gmail address, no wonder I couldn't find it. :) [15:31] wait, why did I think the afternoon standup was now, not half an hour ago [15:31] it hasn't even changed time [15:33] mgz: ah I figured you were skipping the second one as you were at the first one so I rep'd you for it [15:33] mgz: not sure as far as the time slot lack of movement :) [15:33] mgz: to be clear though, the hard requirement is just one of them, two is optional [15:34] rick_h: thanks for repin' [15:41] rick_h: btw, I feel like "verify" is more accurate for what we're doing with the endpoints. To me "validation" means "check against a regex" or some other check of general shape, whereas verification is more "actually try it". [15:41] rick_h: but maybe I'm splitting hairs [15:42] natefinch: I'm find with the verify internally, but I feel like end user input validation is the end user side of things [15:42] natefinch: validate all the user's input, you might validate by verifying the url is reachable...etc [15:44] rick_h: *shrug* ok === freyes__ is now known as freyes [18:21] do we care that the bridge script (based on my current changes) would no longer work for 1.25? [18:21] rick_h: ^^ [18:21] because what I'm doing now makes this a reality [18:22] my expectation is that 1.25 is bug fix only, so we would do the minimal anyway [18:29] frobware: +1 [18:30] rick_h: right now it's a mutiny - characters are being dropped, functions have gone, compatability has been tossed... :) === frankban is now known as frankban|afk [19:38] morning folks [19:39] morning [19:45] morning! [19:53] is the maas API only exposed over http? [19:54] natefinch: or https? [19:54] correct [19:54] and is it only exposed over the standard ports, or is that configurable? [19:55] natefinch: so http/https via apache. [19:55] natefinch: it could be custom if the user monkeys with it [19:56] bleh, ok [19:56] does juju support all that? [19:58] natefinch: you tell me, can you specify the port as part of the url? [19:59] you certainly can.... I presume it just works as long as we don't do anything stupid [20:00] natefinch: I'd assume so [20:06] voidspace, you aournd still? [20:39] menn0: take a look at https://github.com/juju/juju/pull/6622 please? [20:40] menn0: also - do we only want to delete logs for a model when it's migrating? Or should we also do it when it's destroyed? [20:40] babbageclunk: will do in 5 [20:40] babbageclunk: certainly when it's destroyed too [20:41] menn0: Ta, not blocked so no rush. [20:41] menn0: Ok [20:41] babbageclunk: it's an oversight that we're not doing that already [20:42] babbageclunk: the log pruner will catch them eventually (all gone in 3 days), but better to do it up front [21:24] Bug #1645477 opened: Problem deploying openstack-ha with juju deploy [21:33] Bug #1645477 changed: Problem deploying openstack-ha with juju deploy [21:36] Bug #1645477 opened: Problem deploying openstack-ha with juju deploy [21:45] GAAHHHHHH [21:45] man, I can't tell you how much I hate test suites [21:46] natefinch: what's the problem? [21:47] setup test doing things I didn't realize it did, which causes me to waste an hour debugging why my test is passing when it shouldn't [21:56] babbageclunk, I am going to be late [21:56] will ping [22:01] alexisb: no worries [22:05] ok babbageclunk I am joining the HO [22:52] babbageclunk: did you still want me to review that change given that thumper has already approved it? [22:54] menn0: he's approved the removing logs one - I don't think you need to look at it, it's pretty straightforward. [22:54] babbageclunk: I just quickly looked anyway and it LGTM [22:54] menn0: I think you should still take a look at the other one. [22:55] babbageclunk: will do [22:56] menn0: cool thanks [23:16] babbageclunk: done [23:16] menn0: thanks - going through them (and thumper's) now [23:46] perrito666, ping [23:46] thumper, ping