=== beisner- is now known as beisner === meetingology` is now known as meetingology === bradm_ is now known as bradm [08:14] morning all [08:37] any more thoughts on this review? https://github.com/juju/juju/pull/6414 [08:39] morning mgz [08:48] hoenir: I'd be happy to look over as well, but seems you have addressed reviews from axw and katco already? [08:49] hoenir mgz: I will take another pass, but nearly EOD so not until tomorrow [08:55] axw, ok thanks ! [09:36] mgz, hey - we're hitting a juju ssh problem in a charm school at the openstack summit [09:36] aware of any issues? basically the terminal is foobar'ed when you ssh to a ubuntu machine from a windows command prompt. [09:37] jamespage: hm, there have been some issues, let me see what I can find [09:38] jamespage: bug 1468752 [09:38] Bug #1468752: "juju ssh" adds an additional strings to all commands when used on Windows, in interactive mode [09:38] mgz: jamespage: there is this on.. https://bugs.launchpad.net/bugs/1635622 [09:38] Bug #1635622: 'juju ssh ...' fails with Permission denied (publickey), for only one or two machines in a deployment [09:39] ...bug doesn't document a workaround as such, but does suggest spaces before and after the command [09:47] anastasiamac_, ssh key is working ok - but shell is non-functional [09:48] jamespage: is terminal completely borked of does ` echo test ` actually output 'test' as well as errors? [09:48] -of+or [09:49] may also be work considering using `juju run` instead, if possible? [10:38] finally online, one monitor still not working [10:38] ah well [10:39] whats the best way to include a local version of charmhelpers in the wheelhouse? right now i manually put it there and remove the one that gets downloaded in the build process [11:14] SimonKLB: you're better off asking in #juju - there should be people there who can answer your questions (although maybe a bit later in the day when the Americans arise!) [11:16] voidspace: whats the difference between #juju and this channel? i thought this was for the people developing juju / juju charms while #juju was for juju users [11:17] SimonKLB: this channel is used by those developing juju itself. #juju is for those writing charms and using juju. [11:18] SimonKLB: that's how it works in practise. I'm sorry that I don't know the answer to your question. There *maybe* someone here who can answer your question, but not at the moment it seems. [11:18] voidspace: gotcha, so youre not involved with the charmhelpers and/or charm layers? [11:18] SimonKLB: I don't have much experience with them myself, no. [11:19] do you know of anyone awake at this hour that do? id like to have a quick discussion regarding an issue [11:19] rick_h_: do you know who would be a good person to answer SimonKLB's question? [11:21] that, but even more someone to discuss this: https://github.com/juju-solutions/layer-basic/issues/79 [11:23] SimonKLB: when more people are around I will ask again. It might be wise to ask on the mailing list as that gives people time to respond. [11:23] SimonKLB: let me find you a link. [11:24] SimonKLB: one of these people should be able to help: marcoceppi, lazyPower, arosales, jcastro [11:24] SimonKLB: https://lists.ubuntu.com/archives/juju/ [11:24] dimitern: do you have some time to help me with debugging today? [11:25] mgz: sure - what's up? [11:25] dimitern: yea, the only problem is the timezone difference with those guys.. :) [11:25] dimitern: hangout? maas/network fun mostly [11:25] mgz: joining core now [11:52] morning all [11:56] morning perrito666 [12:15] mgz: you got kicked out? [12:25] SimonKLB: https://pythonhosted.org/charmhelpers/getting-started.html#updating-charmhelpers-packages [12:27] rick_h_: thanks! [12:31] rick_h_: do you know if there is going to be any problems using this while also building the charm normally where the wheelhouse is getting populated? [12:32] isnt the upstream version of the charmhelpers going to be installed as well in that case? [12:35] SimonKLB: sorry not sure. Will have to try it out. === meetingology` is now known as meetingology [13:01] rick_h_: 1:1 time [13:02] dooferlad: omw [13:47] Bug #1589680 changed: Upgrading to cloud-archive:mitaka breaks lxc creation [13:52] dimitern: so on that bug, you're doing an upgrade on the system so doing that install before bootstrap doesn't help? [13:52] dimitern: if you install post-bootstrap is there anything that could be done to get it to that working state? [13:52] rick_h_: it's not an upgrade per se [13:53] rick_h_: it won't work with xenial containers to begin with, without upgrading [13:53] dimitern: let's chat after the standup on it please [13:53] rick_h_: ok [13:56] Bug #1560487 opened: local provider fails to create lxc container from template [14:01] kadams54: voidspace ping for standup [14:01] oops, not kadams54, katco who, isn't here so tab complete fail [14:01] lOl [14:01] I was really befuddled [14:01] kadams54: :P you know you want to join us [14:02] Guilty as charged [14:02] rick_h_: omw [14:11] rick_h_: get SimonKLB all set? [14:13] arosales: yea talked to lazyPower :) [14:13] SimonKLB: great, that lazyPower rocks [14:14] well look at that, this code had a ton of nilptr derefs waiting to happen [14:38] perrito666: obviously the answer is not to use pointers [14:39] natefinch: wish it was so easy [14:39] what, you just delete the little *, right? === kwmonroe_ is now known as kwmonroe [15:40] * rick_h_ lunches [15:52] babbageclunk, ping [16:23] hi all, I desperately need a review on https://github.com/juju/juju/pull/6490 its a bit tricky sorry [16:26] perrito666: oh man, dropping the need for our copied dependencies is awesome [16:26] natefinch: heh sadly that is the last tag that will support juju until we move to 1.7 [16:27] perrito666: wow, they require 1.7? [16:27] natefinch: next tag uses Context [16:27] from context [16:27] ahh, crappage [16:28] We could revert it back to golang.org/x/context but maintaining such parallel code is very unwise [16:30] you could file an issue and maybe they'd put it back to using that one [16:35] perrito666, if we are updating a dep we need a +1 from the tech board [16:36] alexisb: technically we had that dep already [16:36] the juju version was special because? [16:36] alexisb: compatible with go 1.2 [16:37] o yuk, yeah ok I remember that [16:37] ok we are good then thank you [16:37] np, sorry for making the assumption there without asking [17:05] alexisb: pong, sorry - missed your message! [17:05] babbageclunk, no worries, was just going to touch base with you, otp atm [17:07] alexisb: ok, give me a yell when you're free [17:07] brb errand === frankban_ is now known as frankban|afk [17:15] alexisb: actually, pausing for lunch - will ping when back. [17:17] k === ahasenac` is now known as ahasenack === ahasenack is now known as Guest91762 [17:28] i will probably miss the mobility, but for now: wow it is so nice to be back on a desktop with 8 fast cores and 16GB ram [17:29] katco: heh yea somtimes forget what real hardware can do [17:30] i definitely did [17:30] i did not know the pain i was in [17:30] at least i did the experiment :) now i know better [17:59] alexisb: back! [18:00] babbageclunk, still otp [18:00] :) [18:00] alexisb: ok cool [18:17] a lte modem is working better than my isp, this is so sad === ubuntu is now known as Guest22446 [18:43] babbageclunk, I am free when you are [18:44] alexisb: ok, jumping into the hangout [18:44] meet you there [18:45] man, recursive code screws with my head [19:11] * rick_h_ goes to get the boy from school [19:12] rick_h_: aww... POC for openstack: http://pastebin.ubuntu.com/23375669/ === Guest91762 is now known as ahasenack === ahasenack is now known as Guest95404 [19:15] super cool to juju update-clouds today and see the new AWS region show up [19:15] <3 [19:15] \m/ [19:16] lazyPower: while you're here, would love to get your impression of this working example of an interactive add-cloud for juju: http://pastebin.ubuntu.com/23375669/ [19:16] interesting, can this be scripted wtihout the use of `expect`? [19:17] i assume it can with a config.yaml [19:17] lazyPower: nope, but if you want to script you can just pass in a yaml file like we do now [19:17] yep, sgm [19:17] * lazyPower +1's it [19:17] ship it [19:18] cool [19:18] natefinch - i'm probably missing most of the context, the last time i setup an openstack controller was for OIL onboarding. but this looks straight forward. [19:18] and i'm +1 to straight forward data entry forms :) [19:19] lazyPower: before this, the only way to add an openstack or maas or manual provider is to give juju a yaml file you had hand-written that you somehow mangically knew the correct format for [19:19] natefinch: what, you dont? [19:19] natefinch - i recall, and it annoyed me [19:19] <3 [19:28] Anyone have an idea why this detects an error by looking for an error message in the header rather than for an error code? https://github.com/juju/juju/blob/staging/rpc/client.go#L103 [19:33] perrito666, natefinch: ^ [19:36] babbageclunk: I presume both are always set at the same time, not 100% sure though [19:45] natefinch: Well, they weren't in some code I just wrote! :) Took me a while to track down. === alexisb is now known as alexisb-afk [20:23] menn0: ping [20:23] katco: hi [20:23] menn0: hey, hope your 3-day was good :) [20:23] menn0: or 4-day? did you take friday off? [20:23] katco: 4 day! [20:24] menn0: nice :D [20:24] menn0: this is ready for review: https://github.com/juju/juju/pull/6469 [20:24] katco: mostly spent doing hard labour in the garden mind you [20:24] menn0: it always works out that way doesn't it? [20:24] menn0: do you write about your garden anywhere? considering starting one [20:24] katco: yes, but it was kind of rewarding [20:25] katco: looking at that PR now [20:25] katco: github says there's a conflict [20:25] menn0: ta. no rush on that other than i'd like to get it merged before it conflicts :) [20:25] menn0: it's a very simple conflict at the top-level... not substantive [20:25] katco: ok [20:26] menn0: cmd/juju/commands/main.go [20:26] menn0: the review doesn't have to be now; i'm not blocked or anything. just when you have time. before i wake up tomorrow would be nice :) [20:26] katco: ok will do [20:26] menn0: ta! [20:26] menn0: so do you blog about your garden at all? what do you grow? [20:27] katco: no, not at all. I am a complete beginner. this weekend was a lesson in how much I suck at gardening. [20:28] menn0: lol! any words of advice? [20:28] katco: my wife and I have gardened a lot. Our gardening blog is evidently down, which I should fix. But I'll try to get it back up again. [20:28] katco: be careful when using a spade - there can be pipes where you don't expect them :) [20:28] lol [20:29] I cut through an irrigation hose [20:29] natefinch: awesome, thx! [20:29] menn0: rofl [20:29] menn0: what are you starting off with? [20:29] natefinch: and what do you grow? [20:29] katco: if you're going to be in this location through next july, plant garlic, it's super easy [20:29] natefinch: with any luck we will not be. this is a future-looking thing :) [20:29] katco: ok :) [20:30] natefinch: maybe i could do some potted garlic [20:30] katco: garlic grows like a flower bulb - plant in the fall, harvest next july-ish [20:30] natefinch: that's perennial isn't it? [20:30] not really, you uproot the whole plant when you harvest [20:31] natefinch: oh. must have read that wrong... looked into that a month or so ago [20:31] katco: my best piece of advice us to mulch mulch mulch. Makes the garden look nice, keeps soil moister, and keeps the weeds down. [20:31] natefinch: is mulch the same as compost? [20:32] katco: nope. Mulch is the stuff you put on top of the ground to keep weeds from growing. [20:33] natefinch: oh didn't know mulch did that [20:33] * katco is going to challenge menn0 for garden newb status [20:33] katco: for flower beds it's often bark, for garden beds my favorite is salt marsh hay... it's like regular straw, but straw has grass seeds in it, which are obviously not something you want in your garden [20:34] katco: what, like regular updates? :) [20:34] natefinch: do you have to reapply? or once? [20:34] katco: basically just once a season [20:35] natefinch: what do you do with the old mulch? [20:35] katco: it tends to start to break down.. if it hasn't broken down too much, you can reuse it, but often times I just till it back into the dirt. [20:35] natefinch: ah ok. so you don't have to like rake it up or anything [20:36] nah. It's just plant matter... it's actually good for the soil after it breaks down [20:36] cool [20:37] menn0: natefinch: have you seen this? https://farmbot.io/ [20:37] oh oh oh.... if you're planting near a structure - get your soil tested for lead (among other things). Sometimes lead paint can contaminate the ground near buildings. you can get your local state extension office to test your soil - just grab a bag and mail it off. It's like $15. Totally worth it. [20:38] natefinch: wow, does that end up in edibles? [20:38] katco: it definitely can, especially if they're root vegetables. [20:38] wow [20:38] katco: that's pretty awesome [20:38] yeah, haven't seen that machine before. That's amazing [20:39] katco: although this is almost certainly how the matrix begins... the robots start with plants but then move onto people [20:39] natefinch: menn0: it's completely open too. plans, cad, etc. [20:39] menn0: rofl [20:39] menn0: i couldn't have a perfectly fine hobby that gets me away from computers to keep me away from computers. [20:40] haha [20:42] * menn0 counts Github reviews vs reviewboard votes [20:44] morning [20:45] morning thumper, wb [20:45] * thumper is feeling a little sore [20:45] about 15 hours of BJJ over Thursday/Friday/Saturday [20:45] very good though [20:45] thumper: what is BJJ? [20:46] brazillian jiu jitzu [20:46] ahh [20:46] thumper: wow, that's a lot of walloping :) [20:46] mm, so, if the test run fails it marks the PR as failed, if it doesn't will it mark it as success in some form? [20:48] why is our develop branch so screwed? and failing CI? [20:49] katco: menn0 natefinch thumper does any of you know the answer to my question?? [20:50] perrito666: i believe it will [20:50] perrito666: it gets a tick [20:50] I see no tick, Ill try a build by hand [20:54] thumper: i've just been going through some of the recent failures [20:54] and? [20:54] thumper: i'll take this one since I've worked on this recently: https://bugs.launchpad.net/juju/+bug/1632485 [20:54] Bug #1632485: TestAgentConnectionDelaysShutdownWithPing fails [20:54] thumper: some of the windows fails are the usual mongodb failing to come up [20:55] thumper: this looks quite strange and needs attention: http://reports.vapour.ws/releases/4522/job/run-unit-tests-race/attempt/2007 [20:59] menn0: I'm pretty sure I have mentioned more than once that 40 minutes isn't good enough on a small VM [20:59] a success is ~36 minutes [20:59] a 10% variation can cause it to fail [20:59] ish [21:00] thumper: well there's that, but all the ssh unit test failures are interesting as well [21:00] ah [21:00] yeah [21:00] didn't scroll to that [21:11] thumper: this is a new failure that started happening on the 22nd: http://reports.vapour.ws/releases/issue/580bc1a0749a566f2d9176bf [21:11] heh [21:12] yes, using space constraints is dependent on the underlying maas configuration [21:23] thumper: any idea what we need to change to fix the failures? who owns this? [21:24] Bug #1630737 changed: juju should use internal vpc network address space when connected to vpc via vpn [21:26] menn0: so that's something we need to work on the bigger "spaces on aws" picture [21:26] menn0: I just triaged it and we'll make it part of that work we're planning for 17.04 [21:27] rick_h_: but what about the CI failures that are contributing to cursed develop tests now? [21:27] menn0: oh, not sure. [21:28] rick_h_: something spaces related was committed on the 22nd which started breaking a bunch of CI tests: http://reports.vapour.ws/releases/issue/580bc1a0749a566f2d9176bf [21:29] menn0: hmm, but that was on the 22nd and no code landed over the weekend I'm aware of [21:29] menn0: this smells like something else up? [21:31] rick_h_: not sure - maybe whatever it was landed on the 21st, or maybe there's something else going on. i'm just trying to raise visibility. [21:31] menn0: hmm, I wonder if this is related to mgz's new maas related tests he's been working to get going [21:31] menn0: I broke a bunch of things, and should now have fixed them [21:31] menn0: rgr, so that bug isn't related. It's something with aws and a vpn that's going on there. [21:32] mgz: k, if you get a sec can you check with menno on what he's seeing vs expectations? [21:32] * rick_h_ has to get the boy ready for violin lessons and then to join the release call [21:32] menn0: can do hangout quickly if you like? [21:32] mgz: sure [21:33] menn0: I'm in the hangout named core [21:37] * redir goes for late lunch and to run an errand [21:37] bbiab [21:39] thumper: so those spaces failures are all in hand. mgz was making some changes to maas+networking related tests and the failures were teething problems. they shouldn't happen on the next run. [21:40] cool [21:42] Bug #1636307 opened: cannot deploy to network space [21:49] menn0: Hey, are you busy? Can I pick your brains about some more migration stuff? [21:54] thumper, fyi this bug is still preventing blesses on CI: https://bugs.launchpad.net/juju/+bug/1625768 [21:54] Bug #1625768: github.com/juju/juju/state go test timeout [21:54] yes [21:54] I know [21:55] and there isn't really much we can do to fix it [21:55] the state tests just take a long time [21:55] I have pointed this out many times [21:55] until we replace the DB [21:55] or move all the logic out into separate packages [21:55] this will continue to fail [21:56] there is no quick fix for this issue [22:02] thumper, I am on bluejeans [22:02] ack [22:07] bdx: can you poke around in your machine-0.log and see what juju thinks it knows about spaces when it's deploying machines to the wrong places? [22:08] babbageclunk: hi, sorry was AFK [22:08] babbageclunk: can chat now [22:09] mgz: yeah omp [22:11] bdx: thanks! [22:12] Bug #1560487 changed: local provider fails to create lxc container from template [22:16] menn0: Sorry - now I'm AFK - organising burger order but back in 5? [22:16] babbageclunk: np, i'll be here. [22:22] menn0: ok, jump into a-team hangout? [22:22] babbageclunk: ok [22:23] menn0: oh, hang on, that's bluejeans now - core? [22:24] babbageclunk: yep :) [22:34] wallyworld, we are on bluejeans [22:34] I wil be right back [22:35] alexisb: oh, i was in the hangout from the meeting invite [23:09] Oops, burgers just arrived - not going to make it to the standup, sorry! See you next week! [23:16] anastasiamac, ping [23:17] babbageclunk, ping [23:17] veebers, ping [23:17] alexisb: pong o/ [23:17] are you joining the hangout [23:17] standup [23:18] alexisb: ah right, yes I can join, but might need to leave early === mup is now known as 21WAAGPR0 [23:48] veebers: any idea what the deal is with http://juju-ci.vapour.ws:8080/job/github-merge-goose/17/console ? [23:48] veebers: failing because of a missing dep... do they have to be added to the bot machine manually? [23:49] axw: hmm, wallyworld asked balloons the same thing. Let me have a look in a short moment [23:49] veebers: ok, thanks [23:52] axw: veebers: i don't want the branch to land yet as I am pretty sure there's still a launchpad.net/gnuflag dep in there which needs to be culled [23:52] but we can still fix the other dep issue in preparation [23:52] wallyworld: ok. yeah, i did see the bit just above where it grabbed launchpad.net/gnuflag [23:52] sounds good [23:52] secgreoup-delete-all/main.go [23:53] wallyworld: the jenkins job has launchpad.net/gnuflag declared as a dep, that needs to change? [23:54] yes. to the github version [23:54] the job error shows the missing repo [23:54] ack [23:54] wallyworld: when you say you don't want it to land yet, do you want it crippled in some way? [23:55] no, just don't hit $$merge$$ once job is fixed [23:55] once launchpad.net/gnuflag is removed, it won't land anyway :-) [23:56] wallyworld: ack, I'll get the dep updated and you can merge whenever you wish [23:56] awesome, tyvm [23:57] wallyworld: nw, re-building now [23:57] noooo! [23:57] that's what i didn't want :-) [23:57] but it should fail [23:58] wallyworld: carp, I can kill if needed [23:58] hopefully [23:58] we need to update to make consistent use of gnuflags dep [23:58] before landing again [23:58] wallyworld: rats, sorry about that :-\ [23:58] no worries :-) [23:58] wallyworld: it succeeded :-\ [23:59] ah, ok, we'll do another fix up branch [23:59] thumper: when do you usually start?