[00:12] waigani, thumper or davecheney: http://reviews.vapour.ws/r/2721/ [00:12] easy one [00:20] thumper: got 10min to chat? [01:19] rick_h_: I do now [01:36] Bug #1497801 opened: provider/joyent: data races [01:36] Bug #1497802 opened: provider/maas: data races [02:00] Bug #1497807 opened: worker/rsyslog: data race [02:09] Bug #1497809 opened: worker/statushistorypruner: data race [02:09] Bug #1497810 opened: uniter/remotestate: data race [02:12] there are currently 5 data races in juju [02:19] davecheney: is it an improvement? :D [02:19] no, a regression of +5 [02:27] Bug #1497809 changed: worker/statushistorypruner: data race [02:27] Bug #1497810 changed: uniter/remotestate: data race [02:30] Bug #1497809 opened: worker/statushistorypruner: data race [02:31] Bug #1497810 opened: uniter/remotestate: data race [02:37] davecheney: guess that job didn't get made voting then? [02:46] sad trobone [02:58] too many test suites depend on github.com/juju/juju/testing.BaseSuite, when they should be depending on github.com/juju/testing.CleanupSuite [02:58] ^ ie, all the suites that _only_ use PatchValue [03:26] hmmm my merge run is stuck trying to bring up an EC2 instance [03:26] * menn0 cancels it [03:41] * natefinch really just wants to run some shit in a goroutine that reads the db, but in Juju that means a watcher and a worker and an API facade and an api client... [04:06] https://github.com/juju/juju/pull/3340 [04:06] i was able to remove testing.BaseSuite from this package [04:06] davecheney: nice [04:07] ahh, confusing... juju/testing vs juju/juju/testing [04:08] -ETOOMUCHBASESUITE [04:08] +100 [04:40] Bug #1497829 opened: session closed in data source [05:34] * thumper goes to make dinner, will check in later to see william === benonsoftware is now known as bdunn === bdunn is now known as benonsoftware === benonsoftware is now known as MerryChristmas === MerryChristmas is now known as benonsoftware [07:45] fwereade: if you are around? a quick chat [07:46] if you aren't, then enjoy the public holiday :) [07:46] thumper, am here [07:46] I'll be quick [09:02] fwereade: standup? [09:19] frobware: dooferlad: the flag is "source destination check" [09:19] thanks jam [09:21] dooferlad: you can google "ec2 vpc NAT instance" and "Source Destination Check" for more details [09:22] frobware: ^^ would be probably good to understand that discussion, since we're essentially natting for containers [09:23] jam: ack [09:27] jam, dooferlad: for the freeze on the 22nd is there anything we are missing or need to do? [09:31] frobware: have you kept the "15.10 Networking Work Items" up to date? [09:31] I think you and dimiter went over that last week [09:32] frobware: on https://canonical.leankit.com/Boards/View/101652562#workflow-view you can set up a filter using the funnel on the top right. If you use the hilight mode and filter on tag 'phase-1.1-mvp' you can see all cards related to the initial networking support. [09:32] frobware: assuming the tags are correct, that should be everything [09:32] jam: dimiter and I went through that on friday and corrected the dates [09:33] jam: for october there are two items outstanding which is in review [09:34] frobware: there are three cards that aren't in merged. One is docs (assigned to Frank), one is Dimiters code for EC2 (that I can clean up and get someone to review, then land) and one is a low priority code cleanup card. [09:34] jam, dooferlad: and the other item is waiting for James' ci tests to land - so, I believe the October section is up to date. [09:53] frobware, jam: why do we have a doc for that when we have the board? My single source of truth alarm went off. [09:57] dooferlad, jam: predates me but it's easier to see a top-down list vis-a-vis cards on the board which eventually go out of view because they're "done" [09:58] frobware: dooferlad: I would also guess size of scope for each one. I think cards are supposed to be finer grained than an overall summary of things that have been completed. [11:02] jam, want to do the series-metadata thing? if you're comfortable with the phase 1/2 bits I think I am too [11:02] fwereade: I thought you were off for public holiday today [11:03] jam, my concerns are all around the edge cases in the model when we have non-homogeneous units [11:03] jam, yeah, but so was ian and it's important to him :) [11:03] yeah, I'd like to chat, just started a different conversation because I thought you were gone, I'll swing back around in 20 min or so? Is that ok? [11:04] jam, sure, ping me, will try to be near [12:01] anastasiamac: axw ? [12:17] axw: standup [12:21] katco: when you're around, I'd like to meet with you in about 40 min or so if you're up. [13:17] katco: ping if you're around [13:27] jamespage: hey, I just bzr branched ubuntu:juju-mongodb, then I ran apt-get build-dep juju-mongodb but dpkg-buildpackage fails with: dpkg-source: error: can't build with source format '3.0 (quilt)': no upstream tarball found at ../juju-mongodb_2.4.10.orig.tar.{bz2,gz,lzma,xz} [13:27] jamespage: am I missing some step? sinzui you might now too [13:27] perrito666, bzr bd [13:28] (hint use the bzr builddeb pluing) [13:28] jamespage: tx a lot [13:41] perrito666, o/ [13:48] perrito666 anastasiamac: sorry, I just forgot :/ I have set an alarm for tomorrow night === marlinc_ is now known as marlinc [13:58] anybody fancy a fast pr? https://github.com/juju/utils/pull/152 [14:04] lazypower: hi [14:14] Bug #1498010 opened: TestSetsStatusWhenDying

[14:15] well building the mongo deb takes some time :p [14:17] Bug #1498010 changed: TestSetsStatusWhenDying

[14:22] jam: hey around now [14:22] jam: still there? [14:23] Bug #1498010 opened: TestSetsStatusWhenDying

[14:33] jamespage: a bit more help :p at the end it tries to sign the package and fails because I lack the secret keys, any way to go around it? [14:53] katco: I'm stopping by IRC from time to time in case you are around today [14:53] jam: i'm here! :) [14:54] katco: I'm joining https://plus.google.com/hangouts/_/canonical.com/john-meinel-kat [15:14] perrito666, "-- -us -uc" [15:16] tx again [15:27] man, second time we've been creating a service with a default value only to set the value correctly immediately afterward... where the only change to do it atomically is to pass in the value and use it at service creation time. [15:38] katco: fix committed on the 1.25 forward port of the io timeout bug btw. [15:38] natefinch: awesome! make sure LP is updated [15:39] katco: yep. [15:41] gah.... I hate it when people use unnecessary pointers [15:55] o/ Morning, who can i poke about the juju storage provider and loopback storage? [16:15] lazypower: wallyworld, axw, anastasiamac. i actually wrote the loopback provider, but don't have time this morning [16:22] lazypower, what is it that you need? [16:23] is it urgent? [16:32] alexisb, a CPP partner is running into an issue with loopback storage provider, http://paste.ubuntu.com/12515220/ [16:32] it appears its trying to do whats right, but encounters an error and just leaves itself cycling in a broken state. [16:32] lazypower, can you open a bug [16:32] surely [16:33] and I can get eyes on it this afternoon [16:33] lazypower, thanks [16:37] https://bugs.launchpad.net/juju-core/+bug/1498081 [16:37] Bug #1498081: loopback storage provider not completing successfully

[16:38] Thanks for taking a look alexisb [16:38] thanks for the bug [16:45] Bug #1498081 opened: loopback storage provider not completing successfully

[16:45] Bug #1498084 opened: TestManageEnvironRunsCharmRevisionUpdater fails

[16:45] Bug #1498086 opened: TestImplicitRelationNoHooks fails

[17:03] Bug #1498094 opened: TestCommitHook fails

[18:16] katco: you around? [18:16] natefinch: yep [18:17] katco: I think I must be doing something dumb with registering my facade, because i keep getting an error from the API that is unknown object type "UnitAssigner" [18:17] when I clearly do register a facade with the UnitAssigner name [18:18] natefinch: the rpc server uses reflection to check that it conforms to a type signature [18:18] natefinch: so the errors are at runtime [18:18] natefinch: make sure you have the correct methods defined with the exact type signatures expected [18:19] natefinch: does that help, or just stating the obvious? [18:19] katco: both ;) Often, stating the obvious is helpful :) [18:23] lol... if you print something out at the wrong time from the client, it won't bootstrap [18:27] ERROR failed to bootstrap environment: cannot upload bootstrap tools: invalid version "2015-09-21 18:22:51 ERROR juju.apiserver.unitassigner unitassigner.go:15 Registering UnitAssigner\n1.26-alpha1.1-vivid-amd64" printed by jujud [18:41] I bet I know what it is... the version. I bet the default version is 0 for the server and 1 for the client, or vice versa, and so they're not matching up\ [18:41] (version of the facade) [18:42] yep.... client defaults to v0, server defaults to v1. Awesome. [18:45] er, I guess the server doesn't default, but most things are at v1, so I thought v1 would be the default. [18:47] man, would have been nice if the log message included the fact that it was the specific version of the facade that was missing, not the name [20:03] natefinch: hey, need a status update on bug 1497312 for the release meeting. almost done? [20:03] Bug #1497312: make assignment of units to machines use a worker

[20:04] katco: not really... keep running into problems. It's not going to land today unless a miracle happens. Sorry... lots of little things keep cropping up that are hard to debug. [20:36] Bug #1498175 opened: lxcBrokerSuite.* fails on centos

[22:42] perrito666: ping [22:46] thumper: this is not to boast about the rugby game isn't it? [22:46] perrito666: no... [22:47] perrito666: would like to talk work stuff [22:47] if you ahve some time [22:47] sure, need a hangout? [22:50] thumper: ? [22:50] yeah, just otp with alexisb right now [22:50] after that? [22:51] sure, just ping me over here [23:31] thumper: didn't forget about me, right? [23:32] perrito666: nope, ready now [23:32] * thumper creates hangout [23:32] perrito666: https://plus.google.com/hangouts/_/canonical.com/status [23:55] thumper: well that was an awful lot of info to lift from the back of my head, hope I didn't lie too much to you :p if you catch me in the other side of the day I might be more lucid