[00:25] katco: I'm on now [00:26] axw: howdy. we can talk in the standup, not urgent [00:27] okey dokey [00:29] ericsnow: do you still need a review? [00:29] I see Dimiter LGTMd [00:29] axw: nope [00:29] cool [00:29] axw: exactly :) [00:33] menn0: ok... [00:33] menn0: I'd prefer if we fixed the problem [00:33] menn0: perhaps axw could help point in the right direction [00:34] thumper: I'm going to have a look at the same time [00:34] see if I can see why azure's special [00:34] axw: ok, cheers [00:37] thumper: yep i've been talking to axw [00:38] thumper, axw: well anyway I have a branch ready with Ian's changes reverted. unit tests pass and bootstrap on azure works [00:38] so we have that if we end up going that way [00:38] ok cheers [00:38] axw: I've added relevant details to the ticket (including the revs in question) [00:39] menn0: thanks [00:39] * menn0 is desperate for lunch... back soon [00:44] I have nfi how it was ever meant to work, it's using juju-apiserver as the hostname to verify in the client, but the server doesn't have that in its list of host names ... [00:45] if we change the client, it'll reject old servers ... so we'll have to change the certupdater to add juju-apiserver to its list of names [01:00] menn0 thumper: I've got a one-liner fix, will propose after I add a test [01:00] axw: awesome! [01:09] thumper menn0: I think the reason why this only affects Azure is because it doesn't record public IP addresses, only a hostname, and the verification step is skipped if connecting to an IP address [01:21] axw: a one line fix sounds much better than reverting the change [01:28] menn0: RB hook seems to have gone MIA. mind reviewing here? https://github.com/juju/juju/pull/1287 [01:29] axw: looking now [01:33] axw: looks good. [01:33] axw: just one minor comment [01:33] hmm... wondering why it picked up waigani's branch but not axw's [01:33] menn0: thanks [01:33] axw: I agree with menn0 [01:34] axw: and (stupid question) you've tested this with azure? [01:34] menn0: yes :) [01:34] axw: and some other providers? [01:34] nope, I will do before I land [01:34] davecheney, are you awake? [01:34] axw: cool. [01:35] axw: thanks for picking up where I left off and sorting it out [01:35] really glad that the whole change didn't need to get pulled [01:36] mattyw: no [01:36] menn0: no worries [01:36] davecheney, no problem - sleep well [01:37] thumper, menn0: when you have a moment, first pass: http://reviews.vapour.ws/r/602/diff/# [01:37] k [01:39] davecheney, lucky for you I've answered my own question [01:55] waigani: done [01:57] mattyw: did you still want to chat? I have to duck out to do some errands. you'll be around for a while right? [01:57] thumper, menn0: thanks [01:58] I'm just waiting for the bot to be unblocked [01:58] so I can land my branches [02:00] this is a grunty review: https://github.com/juju/juju/pull/1289 [02:00] * thumper waits for RB [02:01] http://reviews.vapour.ws/r/603/ [02:01] anyone bored? [02:07] oh... [02:07] it seems that github is much better at determining the diff than review board [02:07] github has the files moved and edited, whereas RB just thinks they are new and deleted [02:11] axw: tests failed [02:13] thumper: flaky CI [02:13] retrying [02:13] * thumper sighs [02:13] ack [02:15] menn0, I'll be around for a while, whenever is good [02:58] thumper: looks awesome :_) [02:58] thumper: just block remove tests r lost?.. [02:58] anastasiamac: I'm assuming you are meaning the branch :) [02:58] * anastasiamac rools eyes [02:58] anastasiamac: probably... [02:58] rolls even [02:58] anastasiamac: I remember adding one, but that must have been add [02:59] anastasiamac: add a comment and I'll add in the remove one too [02:59] * thumper is now doing environment set/get/unset [02:59] then remove [02:59] well, destroy-environment [02:59] thumper: i have added comments [02:59] thumper: looking forward to new ones [02:59] in preparation for "juju environment share/unshare/create" which will be behind a feature flag [02:59] thumper: supercommands r very neat [03:00] assuming I ever get to land anything [03:00] anastasiamac: ta [03:00] thumper: a pleasure - enjoy === kadams54-away is now known as kadams54 [03:02] anastasiamac: re-declaring table tests [03:02] anastasiamac: we have moved to declaring them with the tests [03:02] inside the test func itself [03:02] that way you don't have types or arrays lying around at package scope [03:02] anastasiamac: I think you just get used to reading it all together [03:03] anastasiamac: when I first started they were very separate, but it meant that the things being tested and the test that tested them were often far apart [03:03] thumper: hmm. it'd be great to have a separate setupTestData to generate tables [03:03] why? [03:03] thumper: otherwise tests r looking sooo unnecesarily veerbose.. [03:04] thumper: verbose even [03:04] you may just have to trust me on this one [03:04] thumper: i have done them this way when i started [03:04] but I can certainly point to poor examples of both [03:04] thumper: but eric pulled me on this [03:04] * thumper will slap eric [03:05] thumper: eric is good - no slapping [03:05] ericsnow: you still around [03:05] thumper: another way would be to use shims [03:05] anastasiamac: as long as the body of the for block isn't too convoluted (and a few are), having it all together is better than the alternative [03:06] * anastasiamac really like shims [03:24] is nz not in asian cup?! [03:33] mattyw: ping? [03:33] menn0, pong [03:33] mattyw: do you want to have that chat? [03:33] menn0, sure - is irc ok? [03:34] mattyw: yep [03:41] anastasiamac: for what? [03:41] thumper: soccer :-) what other sport is there? [03:42] anastasiamac: I have a funny link for you [03:42] SFW [03:42] * thumper works out how to extract from facebook [03:43] ah noo... [03:43] seems like it has been removed [03:45] * anastasiamac waiting for thumper to extract a link... [03:45] https://www.youtube.com/watch?v=8F9jXYOH2c0 had to find it [03:45] the person who originally shared it on facebook removed it [03:49] thumper: oh yeah, i've seen it [03:49] thumper: didnt like it :( [03:49] * thumper is watching it again [03:49] and laughing [03:49] sorry [03:52] thumper: it's k - have ur moment :) [03:52] third time I have watched that today, and it still makes me laugh [03:52] Scott Sterling!!! [03:54] thumper: on bug 1400358 now [03:54] Bug #1400358: TestLoginsDuringUpgrade broken on i386

[03:54] menn0: thanks [04:02] thumper, axw : hmmm looks like that's also wallyworld's certupdater branch... the precise i386 unit test run has failed 100% of the time since that landed. [04:03] thumper, axw : it fails intermittently for me [04:03] hmm [04:04] fun [04:06] closed explicitly smells like a mongo connection being left open for a long time [04:06] thumper, axw: I think I see the problem in the logs. the test that's failing is bringing up an entire machine agent and when the certupdater worker regenerates certs it kills the API worker which drops the connection used by the test and causes it to fail [04:07] thumper, axw: I'll see if I can make the test do it's thing in a different way. [04:07] haha [04:07] damn [04:07] ah heh [04:07] thumper, axw: i've never been that happy with that test and a few of the others on the same suite [04:08] menn0: feel free to hack and refactor [04:08] thumper: well I wrote this test in the first place :) [04:08] ha === urulama_ is now known as urulama [08:20] morning [08:48] fwereade: can you please take a look at this later? http://reviews.vapour.ws/r/589/diff/# [08:48] axw, thanks, will do [08:48] fwereade: one thing that feels a bit iffy to me is Source; it's there primarily because of shared storage [08:49] axw, ok, I'll keep that in mind [08:49] thanks [08:50] fwereade: I did also respond to your comments and make some updates to https://github.com/juju/charm/pull/77, but that's less important right now [08:50] the latter will feed the former [08:51] axw, yeah, sorry I didn't do that properly, I got grumpy about the bson requirement and then got distracted by something and haven't got back to it [08:51] fwereade: no worries [08:58] fwereade, ping? [09:12] jam1, ping? [09:12] hey mattyw, just otp now [09:16] dimitern: when you get a chance, this has been updated http://reviews.vapour.ws/r/596/ [09:17] dimitern: the race condition is tested [09:17] dimitern: it occurs to me that the race is tested by mocking out the rand.Int63n call - so it doesn't actually *need* to be tested with concurrent goroutines [09:17] dimitern: testing sequential calls would still be the same test [09:17] dimitern: and that would simplify the test a great deal [09:17] dimitern: let me know what you think [09:18] dimitern: I'm on ec2 Subnets - so no great hurry [09:58] voidspace, I'm back just now -- I'll have a look after standup [10:00] dimitern: cool [10:00] dimitern: voidspace: standup? [10:01] jam1, omw [10:07] jam1: dimitern: TheMue: I got booted out [10:08] jam1: dimitern: TheMue: probably need to re-auth [10:08] sorry [10:08] I was done anyway [10:08] voidspace: then good riddance anyway :) [10:09] hah [10:09] it won't let me login [10:09] :-/ [10:09] voidspace: sorry, I didn't want to have to tell you this way... :) [10:09] jam1: :-) [10:34] fwereade: big storm on the way here... if i'll lose internet, I am y not b able to talk to u [10:34] s/y/ [10:35] anastasiamac, if you think that's likely, we could talk sooner? I'd like a *bit* of time to tie up some loose ends with what I'm doing though... [10:37] fwereade: k. let me know when u r ready ;-) [10:50] morning all [10:51] voidspace, reviewed [10:51] morning perrito666 [11:09] dimitern: thaks [11:09] *thanks even [11:11] :) [11:20] fwereade: it looks like z storm passed, internet flickered but stayed... I can make allocate time - please do not rush! [11:21] anastasiamac_: zombie invasion? [11:21] perrito666: lol ;-) [11:21] perrito666: with lighting bolts! [11:22] perrito666: and hail :-0 [11:23] perrito666: we maybe not in canada (http://news.nationalpost.com/2013/02/13/canada-will-never-be-a-safe-haven-for-zombies-foreign-minister-john-baird-tells-house-of-commons/) [11:23] perrito666: but we r still pretty hard to invade -.- [11:23] lol [11:34] dimitern: fixed those minor issues and merging [11:35] voidspace, cheers [11:50] anastasiamac_, ha, thanks, it turns out I'm *just* coming up to done now anyway :/ [11:52] fwereade: perfect :-) tyvm for ur time! I'll c u in a few :-) [13:31] GAAAAH if err == nil { return err } [13:39] meh some diffs are so hard to follow [13:39] fwereade: well, ugly but functional :p [14:06] dimitern: could you add your blessing to http://reviews.vapour.ws/r/601/ ? :p it is a 4 line diff and I dont have superpowers [14:07] perrito666, sure, will have a look in a bit [14:07] tx [14:09] I "ship it!"ed it [14:11] perrito666, ship it [14:11] we all shipit'ed now its wwitzel3's problem [14:11] :p [14:12] I just didn't feel like having him wait 12hs until my mentor was back online [14:57] Gna, provider seems to have a DNS problem. Sometimes names are directly resolved, sometimes I need a number of retries. :( [15:08] natefinch: standup? [16:02] ericsnow: ping [16:02] wwitzel3: ready? [16:02] ericsnow: yep [16:56] ericsnow: 15 conflicts only, not bad [16:56] bbl [16:56] perrito666: sweet! :) [17:22] back from purgatory [17:38] katco: turns out, I'm in the minority in thinking of Missouri as "The South": http://fivethirtyeight.com/datalab/which-states-are-in-the-south/ [17:39] lol [17:39] natefinch: well, i narrow my claim to St. Louis [17:39] there are some pretty "southern" areas of MO [17:39] but STL is pretty metropolitan/urban [17:40] katco: heh.... what's funny is the people claiming Pennsylvania is the south... must be all those Mainers ... "Ayuh, that's pretty fa' south" [17:40] natefinch: lol PA? wow [17:40] right? [17:40] there should just be some latitudinal line that defines it and we should leave it at that lol [17:41] katco: what, like some sort of mason-dixon line? http://en.wikipedia.org/wiki/Mason%E2%80%93Dixon_line [17:42] http://en.wikipedia.org/wiki/Mason%E2%80%93Dixon_line#Symbolism [17:42] natefinch: well, maybe a bit more scientific :) [17:43] natefinch: like subdivide the country into two mostly-equal halves. below, south. above, north. [17:43] katco: anything that separates me from Kentucky and West Virginia is good in my book. [17:43] haha [17:43] natefinch: i can't tell you the angst i feel when people lump STL into that category [17:43] natefinch: because mostly it's like any other largish city [17:44] except out downtown really sucks. [17:44] "downtown" [17:44] no one lives downtown [17:44] or very few [17:44] katco: right... pretty much all cities are fairly progressive... [17:45] katco: yeah, downtown-downtown Boston is very expensive to live in, so not a lot of people do, but there's some subsections of Boston that are really crummy. [17:46] natefinch: well, that implies it's desirable to live in downtown boston. our downtown has very very little residential. and like 1 supermarket. [17:46] natefinch: conversely, what st. louisans define as "St. Louis" is *huge*, so there are some really nice "sub cities" of St. Louis [17:47] *nod* People like downtown Boston because you're close to work and theaters and night life. [17:48] natefinch: e.g.: clayton, a largish "sub city"'s skyline: http://farm3.static.flickr.com/2191/2400264760_711ff5fb2d.jpg [17:48] natefinch: and stl's: http://stormhighway.com/stlouisphotos/skyline/east-view-clear-sky-panorama-e-1330pan.php [17:48] natefinch: claytons is about as big, and it's a much nicer city [17:48] *nod* [17:49] i wish we would have had more time when we were in boston to explore [17:50] katco: yeah, that's one thing I dislike about the sprints... not much time to explore. And the problem with you guys being in Lexington is that it's just so far from Boston... at best you're a 30-40 minute car ride without traffic. [17:50] yeah [17:50] i got kind of a nice driving tour on the way back to the airport [17:50] too bad the driver was a racist homophobe [17:53] ouch, sorry :( I guess those sorts live everywhere.... someday they'll all have died off, and the world will be a better place. [17:55] natefinch: https://plus.google.com/u/0/100662126766165980060/posts/B2mzGaPVbfM [18:10] natefinch: you have a minute? [18:11] natefinch: wwitzel3 and I are on moonstone [18:16] ericsnow: brt === kadams54_ is now known as kadams54-away === kadams54-away is now known as kadams54_ [19:44] fwereade: don't suppose you're around? [19:45] natefinch, kinda [19:48] fwereade: I just noticed that environs.EnvironProvider.Validate returns a (presumably modified) config.Config.... so no matter what changes I make to the provider's own configuration struct... they all need to be encapsulated in the config's map[string]interface({} thingy [19:49] natefinch, yeah, sorry, I evidently didn't make that wrinkle clear enough last time we chatted [19:49] natefinch, whatever you do needs to be trivially convertibleback to an m[s]i [19:49] gah [19:50] natefinch, it's not intrinsically enough to sink the idea but it's a complication for sure [19:51] natefinch, OTOH it's just one extra method on the struct type really [19:51] natefinch, might easily turn out to be worth the effort [19:51] fwereade: yeah, I'll see how it looks [19:52] fwereade: what confused me was that we build up this valid environConfig struct.... but then we throw it away and just return the modified config.Config: http://bazaar.launchpad.net/~fwereade/juju-core/provider-skeleton/view/head:/provider/skeleton/provider.go#L71 [19:52] sinzui: the landing bot is out of disk space, what do we need to do? [19:52] thumper: pay for a bigger disk :) [19:53] thumper: and/or fix whatever is leaking files [19:53] hmm... [19:53] euca-create-tags: error (InvalidInstanceID.NotFound): The instance ID 'i-79999a93' does not exist [19:53] Test failures, reporting on proposal [19:53] thumper, I have no experience with the landing bot, but I am very good at deleting everything in /tmp and zillions of tarballs left behind [19:53] heh [19:54] natefinch, yeah, the main point there is to make sure there OAOO validation path that's shared by both the config-level stuff and the env.SetConfig stuff [19:54] sinzui: why does the landing bot think it is something else? [19:54] natefinch, so long as that property stays I'm not bothered what path we take to get it [19:54] natefinch, within reason [19:54] sinzui: looking at http://juju-ci.vapour.ws:8080/job/github-merge-juju/1573/console [19:54] fwereade: I always try to stay within reason :) (within reason) ;) [19:54] sinzui: who does look after the landing bot? [19:54] natefinch, haha [19:55] rick_h_: you have some understanding of the landing bot yes? [19:55] thumper, ah. well that isn't the merge job. that slave has 23G free [19:55] * sinzui looks [19:55] thumper: mgz maintains the landing bot, no? [19:55] sinzui: the previous attempt failed last night with out of disk space [19:56] I tried again this morning, and got the above error [19:56] thumper, I think that error means ec2 expired the ami and we need a new one [19:56] huh? and why? [19:56] how do we get a new one? [19:57] thumper: yes, wrote the originaly [19:57] yep the juju-core-slave really has 24G free and the most full partition is just at 18% [19:57] * sinzui looks up instance and ami [19:57] rick_h_: seen the error listed here? http://juju-ci.vapour.ws:8080/job/github-merge-juju/1573/console [19:58] thumper: that's not the landing bot but the job that's running and it's config [19:58] ++ /var/lib/jenkins/juju-ci-tools/ec2-run-instance-get-id [19:58] was run and seems it found a non-existant instance on there? [19:58] so just CI stuff... [19:59] thumper: yea, the landing bot doesn't directly touch the cloud, just github api [19:59] ah [19:59] kk [19:59] * thumper is frustrated [19:59] it has been over a week since I have been able to land code, and now the bot is unblocked, I still can't land stuff [20:00] heh, and now hangouts hate me [20:00] "there is a problem with this call, please try again in a few minutes" [20:00] thumper: tried firefox? [20:01] almost feel like checking the whole day in now [20:01] s/checking/chucking/ === kadams54_ is now known as kadams54-away === kadams54-away is now known as kadams54_ [20:11] sinzui: any info about the instance id? [20:17] thumper, still looking...I think the id was hardcoded, damn it [20:26] thumper: hey you reviewed http://reviews.vapour.ws/r/547/ with -1 unless changes where made, I see changes plus I asked the author to file https://bugs.launchpad.net/juju-core/+bug/1400782 which he also offered to address, with all that, do you still -1? [20:26] Bug #1400782: Errors returned by the system are improperly wrapped.

[20:27] aaan, good morning also [20:34] thumper, I have setup the unittests to use a much new ami. you can retry $$merge$$, and we can watch to verify the ami and the instance-type are happy [20:38] dammit, there's an update for chrome (which has been super unstable in utopic)... and apt-get can't download it for some reason [20:38] natefinch: its apt's way to say "stop hitting yourself" === kadams54_ is now known as kadams54-away [20:44] perrito666: ahh, now it's working [20:45] perrito666: if you can call a 54kB/s download "working" [20:46] natefinch: you would not know how often I call that working :p [20:46] natefinch: that usually is your mirror being crappy, just open the conf thinguie and choose a different one [20:53] perrito666: hrm, but it's the chrome PPA, so I don't think I can choose a mirror for that [20:53] ah, you are on your own then :p === kadams54-away is now known as kadams54_ === kadams54_ is now known as kadams54-away [20:59] sinzui: ack [21:00] sinzui: trying again now [21:01] perrito666: I've said shipit now [21:01] perrito666: I'm still not 100% happy, but we can fix it later [21:01] thumper: I am not either, I only said I would go ahead with accepting that solution if the bug was created and addressed [21:02] perrito666: ack [21:28] well that's a first [21:29] refactoring some tests, and i've missed patching a value. brought my system down [21:33] doh [21:41] Does anyone know what causes this error http://pastebin.ubuntu.com/9447169/ or how to resolve? [21:42] thumper: ^ [22:03] thumper: i'm getting trolled on twitter for something canonilca has done [22:03] have we released anothe rproduct ? [22:07] oh, http://www.markshuttleworth.com/archives/1434 [22:09] davecheney: Snappy Ubuntu Core is the new hotness [22:10] davecheney: Do you know how I would resolve this error? http://pastebin.ubuntu.com/9447169/ [22:11] mbruzek: getting trolled on twitter 'cos there is no code [22:11] i'm inclined to agree [22:12] mbruzek: looks like you're deploying on i386 [22:12] davecheney: How would I have done that? [22:13] davecheney: how do I change it? [22:13] dunno [22:14] just trying to decipher that paste failuyre [22:14] basically lxc shat itself [22:14] why ? is not clear [22:16] davecheney: when I do lxc-ls --fancy I see no images. [22:17] but I have an image bootstrapped. There should be some image out there yes? [22:17] mbruzek: you're doing that as root right? if you do it as your own user you won't see any lxc images. [22:18] sudo lxc-ls --fancy yes I am using sudo [22:18] mbruzek: just checking. i've been caught by that before. [22:18] menn0: Yes and thanks for double checking. [22:22] davecheney: yeah [22:23] mbruzek: what does this give you? : sudo ls /var/lib/lxc [22:23] mbruzek: also: lxc-ls --version [22:23] sudo ls /var/lib/lxc [22:23] lxc-monitord.log [22:23] thumper: that xamarin looks cool, thanks [22:24] version 1.0.6 [22:24] urulama_: are you doing mobile dev? [22:25] mbruzek: looking at the error it looks like ubuntu-cloudimg-query wasn't found in the $PATH [22:25] mbruzek: what does "type ubuntu-cloudimg-query" give you [22:26] thumper: yeah, i did ... don't have time currently though [22:27] urulama_: that I get [22:27] mbruzek: if it's missing you'll need to install cloud-image-utils [22:27] thumper: i like doing it directly with ios or android ... but this does look cool. gotta check it out [22:28] menn0: I see one in ~/.cache/ubunu-cloudimg-query [22:28] bash: type: /home/mbruzek/.cache/ubuntu-cloudimg-query: not found [22:29] mbruzek:: on my machine it's in /usr/bin... and it comes from cloud-image-utils [22:29] mbruzek: do you have cloud-image-utils installed [22:29] It does not appear that I do. [22:29] menn0: How was I able to run this before? [22:29] mbruzek: I have no idea... thumper? [22:30] installing cloud-image-utils now [22:30] unless you want me to hold off [22:30] mbruzek: no go ahead [22:31] mbruzek: do you have "juju-local" package installed? [22:31] mbruzek: from the error and from looking at the lxc-ubuntu-cloud script which is used to generate an ubuntu LXC container it really looks like you need ubuntu-cloudimg-query and it comes from cloud-image-utils [22:32] thumper I should. I was running juju local yesterday. juju-local is already the newest version. [22:34] mbruzek, thumper: cloud-image-utils doesn't appear to be directly required by juju-local [22:35] strange, it was not installed on my system... I use local on a daily basis [22:36] not sure what is going on actually [22:36] mbruzek, thumper: the dependency path is: juju-local -> lxc -> lxc-templates -> cloud-image-utils [22:37] mbruzek, thumper: not sure how cloud-image-utils went missing for you... [22:37] lxc-templates is already the newest version. [22:37] I rebootstrapped my environment [22:38] if you do "apt-cache showpkg lxc-templates" you'll see it needs cloud-image-utils [22:38] but somehow it was uninstalled on your system [22:38] menn0: I do apt-get upgrade every morning. [22:39] mbruzek: yeah, this is pretty strange [22:41] mbruzek: with the package installed do things work? [22:42] menn0: Yes it appears that solved my problem. [22:42] many thanks menn0 and thumper! [22:43] mbruzek: please let us know if it happens again. [22:45] menn0: It was reproducible only today for me. I did several destroy-environment, bootstrap with the same error. [22:45] but now it looks good, thanks for working with me on this. [22:45] https://code.launchpad.net/snappy-ubuntu [22:45] sad [23:18]