[00:27] wallyworld: btw, plz tag with "eda" whichever bug u decide is primary for "application in error, can't destroy model" to allow QA the ooportunity to add functional test for this scenario [00:27] opportunity even :D [00:49] wallyworld: also in addition to bug # 1654928, there is possibly related bug # 1661930 [01:06] wallyworld: patience is not on this monday :) I have updated the bugs as per ^^ [03:46] wallyworld: can you please review https://github.com/juju/juju/pull/6924 [04:56] wallyworld: ping? did you see my message? [05:56] axw: sorry, missed it, i had my sound turned down so missed the ping [05:57] wtf, i already reviewed that pr and hit approve [05:58] must have timed out, it does that when i bootstrap on aws and it floods by connection uploading agents [05:59] wallyworld: thanks [06:01] anastasiamac: bug 1465317 isn't critical, the 2.1 binary works fine with current windows/macos and we will have released 2.2 before win 11 or macos 10.13 are close to being released. it's a fairly large fix so should be targetted at 2.2 [06:01] Bug #1465317: ubutu devel macos win: panic: osVersion reported an error: Could not determine series

Fix>

[06:02] wallyworld: sure. plz adjust [06:14] * wallyworld wishes bug reporters would assign bugs to the correct series [06:19] wallyworld: if u r looking for bugs to do, this one is awesome - bug 1654116 [06:19] Bug #1654116: Attempts to write leadership settings when not the leader during relation-changed hooks [06:19]

[06:20] tis ok, i've found one === frankban|afk is now known as frankban [08:50] wallyworld: did u test ur fix in https://github.com/juju/juju/pull/6906 for bug 1643430 with subordinate in error? [08:50] Bug #1643430: Unassigned units in an error state cannot be removed [08:50] wallyworld: bug 1662018 seems to be in the same area [08:50] Bug #1662018: Cannot remove a deployed subordinate charm in a failed hook status [09:00] Bug #1661576 changed: show-status-log formatting contains unwanted pad lines [09:00] Bug #1661624 changed: show-status-log bogus timeline [09:25] anastasiamac: i didn't test with subordiante, fix should be same [09:52] wallyworld: plz either comment on the bug with ur assertive conclusion or add a test \o/ thnx! [10:37] anastasiamac: hiya [10:44] here's a backport of the api dial spurious message fix to juju 2.1: https://github.com/juju/juju/pull/6926 [10:44] reviews appreciated [10:52] morning [10:52] jam: tx for the review on the ipv6 fix [10:52] jam: does cheking on the glibc part of the error seems like enough to you? [10:57] perrito666: "protocol not supported" ? [10:57] "socket: address family not supported by protocol" [10:59] * perrito666 is working at a bar and thinks people look weird at him because he is not using a mac [10:59] perrito666: I'd be much happier if we had a ENOADDRSUPPORT with an official value rather than a random string that someone is going to think could be better worded differently [10:59] perrito666: but I can't think of anything better vs that [11:00] jam: I dont think lxd library is going ot grace us with said code :p [11:18] jam: also I would appreciate your input on how you think this should be tested other than actually invoking lxd (which seems like the wrong thing to do on a test) [11:19] perrito666: can we mock out the thing we are calling SetServerConfig and have it return a custom error? === jamespag` is now known as jamespage [11:26] jam: mot likely, but that woul not test what you want to test which is the fact that the string is still valid [11:45] rogpeppe: hi \o/ m neck-deep in bootstrap failures but there is light \o/ m OCR tomorrow (it'll start here in couple of hrs), if noone else gets to it, I'll look then :D [11:45] rogpeppe: i'll review in about 10hrs? :) [11:45] anastasiamac: np - jam already hit $$merge$$ on it... [11:48] rogpeppe: jam is awesome \o/ [11:49] rogpeppe: IIRC I approved it for 2.2, happy to rubber stamp it for 2.1 [11:49] jam: you added a $$merge$$ comment to https://github.com/juju/juju/pull/6926 which targets 2.1 [11:50] jam: and it's now merged [11:50] rogpeppe: I mean I approved the work when you targetted develop a week or two ago [11:50] jam: i see a comment from an hour ago [11:50] rogpeppe: and thus I approved the one you proposed today [11:50] jam: ah, ok great, i thought perhaps you hadn't realised what you were doing [11:51] well, I generally don't :). but this time I did' [11:51] jam: i'd read the future tense into "happy to rubber stamp it for 2.1" but it was actually the past tense :) [11:51] jam: :) [11:52] I left off the "was" and "is" is a perfectly acceptable word to put in ther. [11:52] rogpeppe: we tend to only approve one copy. thus back-ports and forward-ports r good to self-approve unless there r vast differences with original :D [11:53] anastasiamac: ah, ok. i thought it's worth getting approval that it's worth backporting though. [11:54] jam: ptal https://github.com/juju/juju/pull/6916 ill add a pr to 2.1 too once this one is lgtmed [12:10] Hi, I'm tring to direct the juju client to insepct the index2.json but I'm keep getting this error in bootstrap invalid URL "file:///var/www/html/toolsdir/tools/streams/v1/index.sjson" not found [12:11] in v1 i have just 2 json filex inde2.json and com.ubuntu.juju-devel-tools.json [12:12] also I'm runing juju with like this [12:12] https://paste.ubuntu.com/23940708/ [12:12] any thoughts? [12:14] abentley: could you shed some light on hoenir's problem? I am not sure what --metadata-source expects [12:26] I'm working on the cloud provider and in the process of the actual bootstrap the bootstrap config searches for images and tools in the default ubuntu sources and I specify the --metadata-source and customise it to pick the correct juju agent and image definitions. [12:26] on the cloud provider I mean oracle cloud provider. [12:27] So I don't know why the simplestreams/tools are looking into index.sjson and not looking into index2.json [12:28] also when tring to validate the tools with the juju metadata validate-tools -d /var/www/html/toolsdir --debug I get the following errors/warnings. [12:29] 14:28:05 DEBUG juju.environs.simplestreams simplestreams.go:457 skipping index "file:///var/www/html/toolsdir/tools/streams/v1/index2.json" because of missing information: "content-download" data not found [12:29] 14:28:05 ERROR cmd supercommand.go:458 "content-download" data not found [12:29] why? [12:29] do you have said key? [12:30] https://paste.ubuntu.com/23940769/ index2.json [12:31] and the ubuntu.com one. [12:31] https://paste.ubuntu.com/23940770/ [12:31] perrito666, what key? [12:31] hoenir: I misinterpreted the error [12:32] that error is really not informative :( we should fix that [12:32] Also I generate the tools with this cmd juju-metadata generate-tools --debug --stream devel [12:32] jam: rogpeppe ever heard of it? [12:32] perrito666: heard of what? [12:33] And the documentation about the juju metadata is quite awful on the official site. [12:33] rogpeppe: 14:28:05 DEBUG juju.environs.simplestreams simplestreams.go:457 skipping index "file:///var/www/html/toolsdir/tools/streams/v1/index2.json" because of missing information: "content-download" data not found [12:33] never bumped into it [12:33] hoenir: that is why I asked abentley or eve balloons they are more likely to have tinkered with simplestreams [12:33] brb, need to relocate [12:34] hoenir, perrito666: is it actually allowed to give a file: URL as a metadata source? [12:34] also I'm on the tip of the develop branch [12:35] hoenir: have you done this kind of thing before? [12:35] yeah you can spiecify the directory or you cand specify the url in the --config like --agent-metadata-url [12:35] hoenir: what files are inside that /var/www/html/toolsdir directory? [12:35] hoenir: 'content-download' sounds like the image files, not the tools files [12:35] maybe not, one sec [12:35] Yeah, when I was using the mass in conjunction with juju [12:36] hoenir: http://streams.canonical.com/juju/tools/streams/v1/index2.json clearly has 'content-download'. I was thinking about the image binaries, but agent binaries also need content-download [12:36] (vs AMI, etc) [12:36] Also when you guys will update the provider documentation in the juju/juju repo ? It's pretty outdated. [12:37] I posted above pastes with my jsons. [12:39] hoenir: offhand I'm wondering if your juju client isn't looking for 'devel' tools [12:39] try setting 'agent-stream: devel' [12:39] also the bootstrap --debug file https://pastebin.ubuntu.com/23940814/ [12:39] I see you are passing that in config [12:39] jam: if that is the case, then that error message should really be improved [12:40] After the err returned I os.Exit(1) to stop the execution .. If you have better ways to debug the code/inspect analyze other tha fmt.Printf("%+v\n", that_var), fmt.Println(err), os.Exit(1) please tell me. [12:40] like? [12:40] --config agent-stream: devel [12:40] ? [12:40] like this? [12:40] jam: to something like 'no appropriate tools found' [12:41] hoenir: your bootstrap has --config agent-stream=devel which is the correct syntax for CLI (if you put it in a .yaml file it is : vs =) [12:42] ok I will try now with agent-stream thing. [12:42] but what about the --config development=true ? [12:43] still the same thing. [12:43] "no tools available" [12:46] can anyone point me into some friendly doc on how to generate the tools and images correctly? and what's the best method to point the juju config to it? Also keep in mind this is very experimental and I'm woking to add a new provider. [12:48] Also this is how I generated the images and escaped that "no oracle cloud found in default images error" juju metadata generate-image -a "amd64" -r "us6" --stream "devel" -u "https://api-z52.compute.us6.oraclecloud.com" --debug -i 20160356 [12:48] Is this the correct way? [12:49] this is the tree structure https://pastebin.ubuntu.com/23940870/ [12:50] how my metadata is laid out [12:50] anyone? [12:51] Hoenir I am one of the only ones online now and I don't know the answer sadly [12:52] :( [12:52] If you get a handle of sinzui he will know it [12:53] perrito 666, I don't know why but this should have a simple solution but clearly the lack of documentation really bothers me. [12:53] At this point I don't have any solution to just fmt.Println() os.Exit(1) step line by line and know what is the problem. [12:54] and it's painful but someone gotta do it. [12:54] Also I hate that the golang dosen't have a standardized compile like C has(gdb). [12:55] debugger* [12:55] anyway perrito666 thanks. [12:56] Hoenir I'll make sure to fwd your problem to the right people when they come online [14:07] perrito666: I am not certain what it accepts, either. Probably a path to a directory containing streams is permitted. When I tinker with that, I use the agent-metadata-url config setting. [14:34] Hey I just went through one of these eye exams where they dilate your pupils a lot in one eye so don't expect me to answer chats for an hour or so [14:34] I'll use my good eye to answer emails though [14:47] perrito666: I have these regualarly. in both eyes. It makes getting back from the hospital an interesting exercise... [15:24] frobware: I feel like someone messed up with my white balance and my focus [15:37] morning juju-dev [15:37] morning redir [15:40] * perrito666 is starting to see again [15:40] frobware: how long does this last? [15:41] * perrito666 should have asked before leaving [15:42] morning. [15:45] perrito666: for me around 4 hours. [15:46] perrito666: but also complicated as my pupils are overly dilated due to RP. [15:46] ghaaa 4 hs? [15:46] it has been 2 and still dont see very well so its possible that for me its going to be 3 [15:47] I have the fantastic convo of swollen eye and change in necessary correction === mskalka is now known as mskalka|lunch === mskalka|lunch is now known as mskalka [17:49] Bug #1662272 opened: Agents stuck with "dependency not available" [17:55] Bug #1662272 changed: Agents stuck with "dependency not available" [17:58] Bug #1662272 opened: Agents stuck with "dependency not available" === menn0 is now known as menn0-otp === frankban is now known as frankban|afk [19:39] morning [19:49] morning thumper [19:49] hey [21:15] thumper: do you know who needs to say $$merge$$ on this pr? https://github.com/go-goose/goose/pull/39 [21:15] thumper: can you do it? [21:16] babbageclunk: I'll check [21:17] thumper: clicking through some old PRs I only see axw and bz2 [21:18] I think the person needs write access to the repo [21:18] and I don't have that either [21:18] babbageclunk: I have been told of three sprints this year, the next one next month. [21:18] hmm [21:18] that isn't what I meant to paste [21:18] babbageclunk: https://hangouts.google.com/hangouts/_/canonical.com/morning-chit-chat [21:18] nope [21:18] not that either [21:19] where did all my clipboards go [21:19] lol [21:19] pleased I didn't have anything bad in there [21:19] klajklj83jkhjs [21:19] babbageclunk: https://github.com/orgs/go-goose/people [21:19] babbageclunk: spicer's twitter password [21:19] oh man, now I have to change all my passwords [21:22] wallyworld: Can you merge this please? (or add me to the organisation I guess) https://github.com/go-goose/goose/pull/39 === menn0-otp is now known as menn0 [21:46] do we realy support puma, jaguar, panther? [21:49] redir: ??? [21:49] perrito666: looking at series/supportedseries [22:05] babbageclunk: you should have an invite to goose, see if you can now merge [22:06] wallyworld: awesome, thanks [22:44] wallyworld: did you add babbageclunk to the go-goose org in github? doesn't look like it to me [22:45] https://github.com/orgs/go-goose/people [22:45] thumper: it said an invite was sent [22:45] babbageclunk: did you accept invite? [22:56] thumper: yeah, but I forgot that I needed to make my memebership public. [23:12] babbageclunk: ah... [23:23] thumper: since at the moment you are looking at memory issues, it makes sense to assign bug 1660087 to you from menno - its probably already fixed in an earlier beta; easier to keep such things under the ine umbrella? [23:23] Bug #1660087: Controller and all models unresponsive === mramm_ is now known as mramm === iatrou_ is now known as iatrou === Dmitrii-Sh_ is now known as Dmitrii-Sh === petevg_ is now known as petevg === mpontillo_ is now known as mpontillo === plars_ is now known as plars === cargonza_ is now known as cargonza