[02:58] menn0, waigani_ (or anyone else): if i've bootstrapped and env and then hacked some stuff, how do i "upgrade" the env to the new jujud? [02:58] or is it easier to just destroy and bootstrap again? [03:01] mwhudson: there is juju upgrade-juju --devel [03:01] honestly I jsut guessed thatb [03:01] but there is a flag that builds new tools locally with a minor version one more than what's been deployed [03:02] seems the flag is --upload-tools [03:02] mwhudson: yep --upload-tools is what you want [03:03] mwhudson: remember to "go install" first :) [03:04] heh yes [03:08] heh so http://paste.ubuntu.com/14466151/ is too simplistic [03:09] mwhudson: what happens? [03:10] agent-state-info: 'failed to retrieve the template to clone: lxc container creation [03:10] failed: juju-xenial-lxc-template' [03:12] maybe there is no lxc template for xenial yet [03:23] there isn't, i guess i need to make one [03:23] oh [03:23] no, i think it's probably more that i hacked too hard [03:36] * mwhudson afk for now [04:44] Bug #1532670 opened: Suggest next major version upgrade <2.0>

[04:47] Bug #1532670 changed: Suggest next major version upgrade <2.0>

=== ses is now known as Guest41660 [05:02] Bug #1532670 opened: Suggest next major version upgrade <2.0>

[05:33] If anyone's around, could I get a review? http://reviews.vapour.ws/r/3490/ [05:34] cherylj: i'll look... [05:34] thanks, anastasiamac ! [05:35] and now, off to bed. so tired... === ashipika1 is now known as ashipika === akhavr1 is now known as akhavr [10:03] voidspace, frobware: standup? [10:03] dooferlad: omw [10:03] dooferlad: frobware is ill though === akhavr1 is now known as akhavr === gberginc_ is now known as gberginc [12:01] master is blocked at present, new unit test fails. [12:15] Bug #1532777 opened: undertakerSuite.TestEnvironConfig fails

[13:37] dooferlad, that gomaasapi bug we/I looked at late on Friday was because the test didn't enable static-ipaddresses as part of the capability set. [13:47] dooferlad, I take that back. [13:48] frobware: :-| [15:32] Bug #1532831 opened: upgrade in progress - Juju functionality is limited after bootstrap

[15:51] ericsnow: btw there's a couple blank review comments in the review I lefty... just ignore those. I had intended to delete them, and failed somehow. [15:51] natefinch: np [16:02] Bug #1532841 opened: cannot add charm to storage

[16:02] Bug #1532849 opened: precise-amd64 and trusty-ppc64el unittests do not complete

[16:05] Bug #1532841 changed: cannot add charm to storage

[16:05] Bug #1532849 changed: precise-amd64 and trusty-ppc64el unittests do not complete

[16:05] natefinch: let's touch base after lunch on how the demo's looking [16:11] Bug #1532841 opened: cannot add charm to storage

[16:11] Bug #1532849 opened: precise-amd64 and trusty-ppc64el unittests do not complete

[16:18] katco: roger [16:49] frobware: if we get a bless on your revert branch, is it all reviewed to be merged? (https://github.com/juju/juju/tree/1.25-revert-maas-devices) [16:50] ericsnow, katco: reading the HTTP registry stuff kinda gives me the willies... it's a heck of a lot of code that I can't really determine if it's doing the right thing or not. Maybe it's just my unfamiliarity with the HTTP code we have. [16:50] ericsnow, katco: Gonna give it a ship it, but not sure I won't have missed some subtleties. [16:51] natefinch: i agree and left a comment to that effect [16:51] natefinch: i'd rather not ship that until we have empirically tested code that relies on it [16:51] natefinch: i was also unaware that we were taking that on for the demo; it's not represented in the card. [16:51] natefinch: ericsnow: i would have preferred to work on that later. [16:58] katco, ericsnow: gah, I left comments for the http registration on the upload api patch.... this is what you get for using dependent patches, eric :) [16:58] cherylj, I have a "Ship It!" from dooferlad. [16:58] natefinch: ericsnow: yeah it would be nice to do a diff against that other PR to get just the important bits [16:59] voidspace, want to take a quick look? http://reviews.vapour.ws/r/3474/ [16:59] katco: ah, I must have forgotten to rbt post :( [17:16] dooferlad, ping [17:16] frobware: quick look !? that's a 1500 line diff... [17:17] voidspace, ah, yes, well. [17:17] maybe tomorrow morning :-) [17:37] ericsnow, katco: finished my reviews of eric's code btw [17:38] ericsnow: ptal to http://reviews.vapour.ws/r/3392/ whenever you have a moment [17:39] perrito666: k [17:39] perrito666: might be a little while [17:39] np [18:58] ericsnow: bah, zero value fingerprint is not valid [18:58] natefinch: yeah, ran into that earlier [18:59] ericsnow: I think zero value fingerprint is valid for zero bytes of data [19:00] natefinch: it's because it isn't initialized; GenerateFingerprint(nil) worked as an alternative [19:00] natefinch: but I agree the zero value should be okay [19:03] ericsnow: gah, your external tests make using the tests for debugging impossible :/ [19:04] ericsnow: I guess not impossible... but more roundabout than I'd like [19:06] ericsnow: I assumed the hash for no bytes would be empty as well, but I guess it makes sense that the hash is constant length, so you have to define some hash for zero bytes. [19:07] natefinch: right [19:07] natefinch: however, as zero-value fingerprint should be the same a hash of an empty string [19:07] natefinch: we'll fix that later [19:08] natefinch: or at least discuss if that is meaningful :) [19:09] ericsnow: yeah, I can do the generate(nil) for the demo... but I would like to make the zero value valid. Shouldn't be that hard [19:09] natefinch: agreed [19:17] katco, natefinch: I've refactored that API patch to not need the registration (for now) [19:18] katco, natefinch: also, I found a small bug: http://reviews.vapour.ws/r/3495/ [19:20] ericsnow: cool, ship it [19:27] Bug #1514857 changed: cannot use version.Current (type version.Number) as type version.Binary

[19:29] ericsnow: ship it [19:29] ericsnow: also, updated the demo data to use generateFingerprint(nil) [19:29] natefinch: k [19:30] natefinch: confused... why even set it at all? [19:33] katco: because the default value is invalid and fails validate checks later one [19:33] on [19:33] katco: we were just talking about how we should fix that so the default value is valid for zero data [19:33] natefinch: yeah [19:35] ericsnow: ty for the refactor too [19:35] katco: :) [19:42] Is there someone who can do a review for me? http://reviews.vapour.ws/r/3490/ [19:59] cherylj: lgtm. looks much more robust, kudos :) [20:04] katco: thanks!! [20:15] Bug #1532932 opened: Unable to bootstrap the lxd provider on vivid

[20:16] ericsnow: natefinch: fyi i'm collecting lxd bugs here: https://blueprints.launchpad.net/juju-core/+spec/charmer-experience-lxd-provider [20:16] ericsnow: natefinch: the one mbruzek just opened we already knew about, but weren't tracking anywhere [20:17] katco: How do I add the bug there. [20:18] mbruzek: there's a "link a bug report" link; not sure if you have to own the blueprint for it to show up [20:18] mbruzek: i linked it for you [20:18] katco: OK [20:18] thanks for your help [20:19] mbruzek: ty for trying it out [20:24] Bug #1532932 changed: Unable to bootstrap the lxd provider on vivid

[20:31] ericsnow: oops, wrong channel... [20:31] ericsnow: anyway... looks like tests aren't hooking up the resources code [20:32] natefinch: you have to do registration (see state/resources_test.go) [20:33] Bug #1532932 opened: Unable to bootstrap the lxd provider on vivid

[20:36] natefinch: have you been able to run through the demo yet? [20:40] katco: sorry, no, I was waiting for my stuff to land and then I am working on fixing the tests in that one branch... once that's done I can merge eric's branch and do an end to end test. [20:41] natefinch: np. demo > tests right now though [20:41] katco: right, but I can't land if the tests fail [20:42] natefinch: we can create a binary from a branch that's not in github.com/juju [20:42] katco: right... ok, will do that right now. [20:42] natefinch: cool, hope it goes smooth the first time :) [20:44] katco: it would work better if ericsnow's code compiled ;) [20:45] lol [20:45] lol [20:48] 3 silly little fixes... one left [20:49] ericsnow: state.State doesn't implement GetResource? [20:50] natefinch: correct [20:51] ericsnow: fixed it... [20:52] natefinch: fixed what? [20:52] ericsnow: newResourceHandler needed a couple tweaks to compile [20:52] natefinch: yeah, I fixed those [20:52] ericsnow: well then, push ;) [20:54] ericsnow: nevermind, just re-fetched... only one problem now with cannot use *"github.com/juju/juju/resource/api/client".Client as type "github.com/juju/juju/resource/cmd".UploadClient [20:55] natefinch: yeah, the client stuff is not right yet [20:55] reader vs. readseeker [21:00] ericsnow: is there anything I can do to help? [21:00] natefinch: not really [21:02] ericsnow: demo > tests if that helps [21:02] katco: yep [21:44] natefinch: should be ready (sans *any* test coverage on the client) [21:44] ericsnow: cool [21:45] cherylj, ping [21:47] frobware: what up [21:50] cherylj, are we doing some additional runs with the backed-out PRs for 1.25.2? [21:50] cherylj, I was just trying to understand where we are with this before I EOD. [21:51] frobware: yes, it was supposed to get re-run today. [21:51] cherylj, any news? [21:51] jog, sinzui is the revert-branch currently being tested? [21:51] katco, ericsnow: I have a branch, resource-demo (note: singular resource, sorry) it has all the code from everyone. I also have the charm at github.com/natefinch/starsay [21:51] * frobware wonders if no new is good news... [21:51] natefinch: nice [21:52] natefinch: have you tried running through the demo script yet? [21:52] cherylj: No, in the next hour it will be [21:52] katco, ericsnow: just got it building... doing so now [21:52] natefinch: woot woot [21:52] natefinch: cool [21:53] frobware: while I have you here, did you get the chance to figure out what the AWS Robustness for Spaces was referring to? [21:53] :) [21:53] katco, ericsnow: doh. ERROR failed to upload resource "store-resource": PUT https://localhost:17070/environment/2a00757a-9329-401d-8ee5-4140647864c1/environment/2a00757a-9329-401d-8ee5-4140647864c1/services/starsay/resources/store-resource: request body supplied unexpectedly [21:54] cherylj, machine-dep-engine is still running [21:54] natefinch: so awesome! [21:54] katco, ericsnow: that's after deploy, running juju upload [21:54] natefinch: fixing now [21:55] ericsnow: notably, environment/UUID in there twice, too === Guest3473 is now known as Makyo [21:55] natefinch: indeed [21:56] katco, ericsnow: btw, I updated my demo data at deploy branch so the tests will pass: http://reviews.vapour.ws/r/3484/ [21:57] cherylj, I sent some email dimiter - but we got caught up with demo stuff. he's back on Weds and will follow-up then. apologies for the delay. [21:57] frobware: okay, I'll move it to alpha2 until we can get it figured out :) [21:57] katco, ericsnow: it injects the resource code into a registry now, which is probably more like how we'd want to do it (though I imagine we'll actually want to inject DB ops to keep it transactional)... but it means that if you don't have resources registered, we won't try to use resources. [21:58] ericsnow, katco: I gotta run for a few hours for dinner and bedtime for the kids, but will be back after. [21:59] natefinch: kk ty [21:59] ericsnow: can you pick up where he left off> [21:59] ? [22:00] oh and to potentially save you a few minutes figuring it out, use 'juju deploy ~/src/github.com/natefinch/starsay --series trusty' to deploy starsay locally (so you don't have to put it in a charm repo or special folder) [22:00] (obviously replacing that path with the path where you clone the repo) [22:01] ok, afk now === natefinch is now known as natefinch-afk [22:02] katco: k [22:03] Bug #1527068 changed: MAAS retains child devices' IP addresses when a parent node is released

[22:06] Bug #1527068 opened: MAAS retains child devices' IP addresses when a parent node is released

[22:07] cherylj, bug #1525280 is now for alpha2, so not a cut-off for this weds? [22:08] Bug #1525280: 1.25.1 with maas 1.8: devices dns allocation uses non-unique hostname

[22:08] frobware: isn't that bug no longer relevant when we back out bug #1483879? [22:08] Bug #1483879: MAAS provider: terminate-machine --force or destroy-environment don't DHCP release container IPs

[22:08]

[22:08] oh wait [22:08] cherylj, on master? [22:09] you're talking about master [22:09] yeah [22:09] I just kinda assumed that you guys weren't going to get to it before then [22:09] trying... [22:09] frobware: if you think you can get it in in the next couple days, then we can move it back [22:09] frobware: but really, there are some branches that are trying to land that are holding up alpha1 [22:09] cherylj, ok. I'll do some manual (yuck!) testing tomorrow. [22:10] frobware: if you need any help, let me know [22:11] cherylj, I'm beginning to like the (over the wall) testing that gets done by a feature branch. [22:11] frobware: yeah it's nice. We just need to get more hardware for the QA team since the number of branches we have to test has gone up so much! [22:11] cherylj, arguably, to me at least, seems the most expedient thing we could do. [22:12] frobware: there is hardware that hasn't been set up by IS yet [22:12] I'm trying to find a name of someone who can help hurry that along [22:12] Bug #1527068 changed: MAAS retains child devices' IP addresses when a parent node is released

[22:14] hey natefinch-afk - is your fix for bug 1486553 in master? [22:14] Bug #1486553: i/o timeout errors can cause non-atomic service deploys

[22:15] natefinch-afk: hmm, it's old. Looks like it's there, though [23:00] cherylj: hello, how can I help ? [23:00] hey davechen1y welcome back! [23:00] davechen1y: we're seeing seemingly random test failures on wily and xenial [23:01] davechen1y: sinzui is working on compiling a list of the failures for you to look at [23:01] cherylj: ok, i'll get on it [23:01] thanks davechen1y! [23:02] can you paste links to the failing builds ? [23:02] sinzui: do you have examples you can give davechen1y to look at while you're compiling the list? [23:06] cherylj: Still pasting. [23:30] davechen1y: cherylj: This is what we know, which is not much :( https://pastebin.canonical.com/147376/ [23:31] sinzui: I know this is probably a stupid question but, do you have a moment? [23:45] perrito666: I do have a moment [23:49] sinzui: trying something else so I dont waste my time when I ask you the question [23:49] s/my/your