[00:48] thumper: time for a chat ? [00:48] yep, 2 minutes? [00:52] mkay [01:23] does anyone know the way to bootstrap with --upload tools if i am running trusty but need to bootstrap a precise environment [01:23] because reasons [01:26] going to try juju bootstrap --bootstrap-series=precise --upload-tools [01:41] looks good ! [02:14] thumper: this machine just decided to die [02:14] http://paste.ubuntu.com/14659482/ [02:15] awesome [02:15] just didn't feel like getting an IP address, ¯\_(ツ)_/¯ [02:18] 2016-01-25 01:44:35 ERROR juju.worker runner.go:226 exited "machiner": setting machine addresses: cannot set machine addresses of machine 3: state changing too quickly; try again soon [02:18] fucking great [02:18] who's changing state so quickly ? [02:19] this is the only agent left running ? [02:23] ah [02:23] this normally means an assertion is failing [02:23] poor error message [02:26] i shall raise a bug [02:31] thumper: https://bugs.launchpad.net/juju-core/+bug/1537585 [02:31] Bug #1537585: machine agent failed to register IP addresses, borks agent [02:55] Bug #1537585 opened: machine agent failed to register IP addresses, borks agent [02:55] Bug #1537586 opened: tools version checker spams public release bucket for unreleased tools [03:06] lucky(~/devel/swift-test) % juju status --format short [03:06] - glance/0: (error) [03:06] - keystone/0: 54.253.213.57 (started) [03:06] - mysql/0: 54.206.217.205 (started) [03:06] - swift-proxy/0: 54.79.202.139 (started) [03:06] - swift-proxy/1: 54.79.198.107 (started) 8080/tcp [03:06] - swift-proxy/2: 54.253.216.163 (started) [03:06] - swift-storage-z1/0: 54.206.230.31 (started) [03:06] - swift-storage-z2/0: 54.66.53.57 (started) [03:06] - swift-storage-z3/0: 54.206.156.78 (pending) [03:06] ffs [03:06] failed again [03:06] see glance/0 has no ip address [03:07] ugh [03:07] that blows [03:25] oh, and the you cannot delete the env [03:25] because it goes apeshit that security groups are in use [03:25] and then locks you AWS account because of rate limiting [03:57] juju deploy, http://scorpiondagger.tumblr.com/post/100098695703/check-out-the-good-book [04:29] thumper: that "too fast" error [04:30] is very worrying [04:30] its coming from the txn package [04:30] the environment is idle [04:30] nothign deployed because one of the machiens could not register it's ip [04:30] but the txn package has fucked up copletely [04:30] to be fair [04:30] this was on a precise instance [04:30] and maybe the version of mongo installed is buggered [04:31] but this environment is idle [04:31] there should be nobody adding txn's to the queue, and all the participants should be attemtping to clear the txn queue [04:37] wallyworld: starting on your 9-page diff now... :/ [04:37] sorry ;-( [04:38] it's not as bad as the 17 page one i did for anastasia :-) [04:38] wallyworld: heh :p [04:40] OH MY GOD [04:40] the config-changed hook of swift does the whole apt-get update/upgrade/install dance [04:40] over and over again [04:42] davecheney: the "changing too quickly" error is almost certainly a bad assertion that keeps failing [04:42] * thumper is heading off for the day [04:49] it's coming directly from juju/txn.Run [04:49] ahh [04:49] ok, so it's not that the txn is not comitting [04:49] but it fails ? [04:50] and txn cannot tell the difference between an assertion failure, and a general failure ? [04:50] \o/ [05:04] Bug #1537620 opened: environs/ec2: a stopped machine will cause kill-controller to fail and blow the rate limit [05:43] axw: thanks for review, will fix that method. got a fark load of conflicts with previous work to resolve first, sigh [05:44] wallyworld: nps [06:37] wallyworld: I'm going to assume your changes are just changing env->model, as requested, rather than going through another 7 pages of diff. I see you changed the method name. [06:48] axw: yeah, correct. the method name plus more things like envUUID -> modelUUID etc [06:48] and a file rename or two [06:48] environdestroy.go -> modeldestroy.go [06:49] axw: if i can land this one, plus the latest api alias one, plus possibly 1 more i'll do after soccer tonight, we should be good for a CI run using the 2.0 scripts [07:45] wallyworld: sorry, went afk for a while [07:46] will review the last one now [07:51] wallyworld: shipit [07:52] * axw goes back to debugging quassel on his laptop [09:59] dimitern: dooferlad: frobware: grabbing coffee [09:59] sorry [10:02] oops omw as well [10:35] dimitern: space "name translation" has now landed on maas-spaces, it's easy code to understand but you may want to look at it to check you're happy with the algorithm [10:35] dimitern: very easy to change if you're not (in the discoverspaces worker) [10:38] will do [11:06] voidspace, I'm back... [11:06] frobware: let me grab coffee :-) === natefinch-afk is now known as natefinch [11:13] frobware: right, omw [11:44] Bug #1537717 opened: status format should default to tabular in 2.0 [11:53] Bug #1537731 opened: Repeated log handling getting recursive

[12:02] Bug #1537731 changed: Repeated log handling getting recursive

[12:05] Bug #1537731 opened: Repeated log handling getting recursive

[12:29] Bug #1537740 opened: sudo error on bootstrap and working on units [12:44] Bug #1537740 changed: sudo error on bootstrap and working on units [12:47] morning all [12:50] Bug #1537740 opened: sudo error on bootstrap and working on units [14:11] sinzui: You've seen ian's mail? [14:58] wow lxd provider didnt even flinch at a power outage during a deploy, that is awesome [14:59] perrito666: <3 [15:43] dimitern: frobware: dooferlad: merge master onto maas-spaces again [15:43] dimitern: frobware: dooferlad: http://reviews.vapour.ws/r/3624/ [15:45] voidspace, is this at all different when we tried earlier today? [15:47] frobware: no, just updated and nothing new came in [15:48] frobware: but I omitted to actually turn it into a PR earlier today... [15:48] voidspace, ok. LGTM based on my brief experiment with the merge earlier. However, ... my change into maas-spaces end of last week generated CI based unit test failures. did you get a clean unit test run? [15:49] voidspace, the failures were transient as my next $$merge$$ was OK [15:50] frobware: yep, clean run - and I saw those failures last week [15:50] I've had the same problem before unfortunately :-/ [15:50] voidspace, cheers - looking [15:50] frobware: thanks [15:51] voidspace, locally or always in the CI run? [15:51] voidspace, lgtm === kwmonroe_ is now known as kwmonroe [16:06] frobware: I've only ever seen it on CI run, never locally [16:06] dimitern: thanks [16:06] voidspace, ditto [16:44] blue rock guys, assess_container_networking fails on master and also doesn't clean up security groups neatly, anyone free to have a look? [16:47] mgz: I actually had to google that [16:47] mgz: that would be me? [16:48] ...I actually don't know, you're on that team? [16:48] I thought you were on a different rock colour [16:48] mgz: google says that it could be: sapphire, tanzanite, topaz, zircon, spine, aquamarine, apatite, lapis lazuli, tourmaline and a few others [16:48] ehehhee [16:49] too many blue [16:49] tanzanite is more yellow [16:49] http://www.gemselect.com/gem-info/gems-by-color.php [16:49] natefinch: can you take another look at http://reviews.vapour.ws/r/3599/ ? I took a different approach. [16:50] okay, purple. I do not know my rocks. [16:50] frobware: going to be late to our meeting. Need to pick up daughter. She may be watching for a bit! [16:52] cherylj: nice, really like where that ended up. Ship it. [16:52] thanks, natefinch! [16:53] cherylj: looks good, thanks! [17:52] katco: can I bug you for a small goamz review? github.com/go-amz/amz/pull/65 [17:52] mgz: sure... will be after my next meeting [17:52] mgz: (hour or so) [17:52] ta [17:59] mgz: or you know, it can be a 30 line diff and i can do it quickly :) [17:59] mgz: review up [17:59] :D [19:01] mgz, natefinch - can you guys look at my master change for the x-javascript bug. I couldn't do a cherry-pick since the code's changed, but it's pretty much exactly the same: http://reviews.vapour.ws/r/3627/ [19:16] Bug #1491688 changed: all-machine logging stopped, x509: certificate signed by unknown authority

[19:19] Bug #1491688 opened: all-machine logging stopped, x509: certificate signed by unknown authority

[19:22] Bug #1491688 changed: all-machine logging stopped, x509: certificate signed by unknown authority

[19:25] Bug #1491688 opened: all-machine logging stopped, x509: certificate signed by unknown authority

[19:31] Bug #1491688 changed: all-machine logging stopped, x509: certificate signed by unknown authority

[19:34] Bug #1491688 opened: all-machine logging stopped, x509: certificate signed by unknown authority

[19:38] cherylj: ship it! [19:40] thanks, natefinch! [19:40] Bug #1491688 changed: all-machine logging stopped, x509: certificate signed by unknown authority

[19:40] Bug #1537880 opened: Juju missed a MAAS node transitioning to "deployed", shows it as pending

[20:02] fwereade: you around for a quick question not related to tests? ;) [20:14] Could someone online with knowledge of linux comment on this PR? https://github.com/natefinch/lumberjack/pull/20 Guy wants to set umask to 0002 in one of my tests, because his umask is set to 0022, which makes the test fail. I understand what umask is and everything, but this just doesn't quite sound like the right fix, but I'm not familiar enough with Linux to know if there's a better way, or what.. [20:16] uf, syscall umask, big nono unless you know what you are doing [20:17] natefinch: no, if this fails inside a temp dir you create for this purpose with the defaults of the system it is most likely a missconfiguration on their system [20:19] natefinch: where is maketempdir defined? [20:19] perrito666: https://github.com/natefinch/lumberjack/blob/v2.0/lumberjack_test.go#L635 [20:20] perrito666: maybe that mkdir should be 0770 [20:20] ? [20:20] 700 actually [20:20] perrito666: right, that makes sense [20:20] or even 600 [20:21] * natefinch is from windows, I just keep typing 7s until it works [20:21] natefinch: dont make me go there and hit you with a unix manual [20:21] lol [20:21] I'm mostly joking :) [20:22] perrito666: thanks, I'll set it to 0600 tonight and ask the guy to retry that way. I was pretty sure setting umask was not the answer. [20:24] I guess I could also set my umask to 0022 and test it :) [20:30] natefinch: what was the type you were talking about returning from the API layer now for listing resources? [20:39] katco: still a ResourcesResults, which has a []ResourcesResult per service, but now each ResourcesResult will also have a []UnitResources, which will have the unit tag and a []Resource [20:43] natefinch: do you have a branch up i can look at? [20:44] katco: I will in just a bit... I had to undo a bunch of my work from Friday, and getting something that'll compile is taking a little bit of time. [20:44] natefinch: k, please ping me when it's up [20:44] katco: will do [21:24] katco: it's get-unit-resources, the key change to the api is adding the UnitResources field here: https://github.com/natefinch/juju/blob/get-unit-resources/resource/api/data.go#L53 [21:24] katco: I still have to fix the tests, but that shouldn't be too bad. [21:26] natefinch: ty === urulama is now known as urulama__ [21:43] whoever made params.Entities have a field called Entities of type params.Entity should be ashamed of themselves. [21:44] er []params.Entity for that last bit. gah [21:53] hehehe [21:54] that is a common pattern in params [21:55] it's just awful in struct literals [21:56] and elsewhere: args.Entities.Entities = append(args.Entities.Entities, params.Entity{ [21:57] anyway... back in a bit === natefinch is now known as natefinch-afk [23:04] Bug #1537937 opened: localHTTPSServerSuite no trusty arm64 images

[23:13] Bug #1537937 changed: localHTTPSServerSuite no trusty arm64 images

[23:22] Bug #1537937 opened: localHTTPSServerSuite no trusty arm64 images