=== Ursinha-afk is now known as Ursinha === almaisan-away is now known as al-maisan === Ursinha-afk is now known as Ursinha === Ursinha-afk is now known as Ursinha === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [08:09] mthaddon: hai [08:27] lifeless: can I assist? [08:28] I want to change the prod timeouts. [08:28] and edge [08:28] which ones? [08:29] db timeout and soft timeout [08:30] db timeout being the 20secs and then we oops? I suspect stub'll know where that hides. [08:30] oh, I know where it is. [08:30] where is the prod config branch ? [08:30] oh. sorry, I misunderstood the Q. [08:31] lp:~launchpad-pqm/lp-production-configs/trunk [08:36] https://code.edge.launchpad.net/~lifeless/lp-production-configs/timeouts/+merge/30069 [08:38] i'd grumble at the 'updating diff' message, but that'd be my problem to chase. so I won't. [08:38] spm: ^ I want that to be live on the prod,edge,staging appservers, please. How do we make that happen ? [08:39] lifeless: edge (I'm pretty sure on edge and config updates, but not 100%) and staging will just happen; tho if you want faster, edge can be manually done. prod is a CP. [08:40] spm: its a CP to change the _config_ ? if so fine, tell me what to do. [08:40] spm: I'd like to do this broadly in lockstep so that the oops hourly graphs are comparable [08:42] lifeless: we don't really distinguish between code vs config changes. they all follow much the same process. [08:44] lifeless: so the request will be (after approval/buildbotting etc) wind up here: https://wiki.canonical.com/InformationInfrastructure/OSA/LaunchpadProductionStatus there's a link off that to the process for getting a change thru [08:46] spm: please tell me you're kidding about buildbot for this? [08:47] buildbot doesn't use the production configs AFAIK [08:47] so I'm entirely unconvinced it has any use here [08:49] hrm. well I was speaking to the general case not the specific. I guess we do send all stuff thru BB as a matter of course. If only such that the appropriate branhces wind up in the appropriate spots. [08:50] In the case where a change is sufficiently ZOMGness, we can and do cowboy in. [08:50] well, have a look at it. [09:06] https://bugs.edge.launchpad.net/wadllib/+bug/274074 [09:06] <_mup_> Bug #274074: Missing total_size on collections returned by named operations [09:16] mwhudson, you'll need the LP branch plus http://paste.ubuntu.com/464412/ and a checkout of lp:~salgado/lazr.restful/extension-interfaces on lazr-restful-dev-egg [09:17] salgado: ta [09:33] spm: hi [09:33] spm: I realise its well past your off-time, but we got left half-discussed; If you don't reply I'm going to assume you're having fun and will follow up with mthaddon [09:34] heh, having fun trying to keep codebrowse alive :-) [09:47] spm: ouch [09:49] mthaddon: whats the pqm info for the prod configs branch ? [09:50] lifeless: as in, which branch to submit to? bzr+ssh://bazaar.launchpad.net/~launchpad-pqm/lp-production-configs/trunk [09:53] and the pqm email box ? [09:53] mthaddon: ^ [09:53] lifeless: same as for all the LP branches [09:53] mthaddon: its been many years since I had that setup [09:53] mthaddon: ec2land does it magically [09:54] hmm, might it be pqm at pqm dot ubuntu dot com ? [09:54] lifeless: there's no developer docs on that? I can find out for you, but I suspect it should be documented for developers somewhere [09:54] pqm_email = Launchpad PQM [09:54] lifeless, mthaddon: ^ [09:55] mthaddon: the docs are a little awkward right now :P [09:55] thx - not quite the developer docs I had in mind, but... :) [09:55] StevenK: thanks! [10:04] sinzui, I have a simple branch needs a review and yours could qualify for code and ui... [10:04] sinzui, I've just changed a help link that is failing. Can you help with this? [10:10] stub, could you test/land my flags db patch? [10:10] i ran ec2 test but it seemed to hang [10:10] poolie: ok [10:10] or at least the web page did not show progress [10:10] and i suspect it couldn't authenticate to send mail [10:20] can we have lights please ? [10:27] salgado: ICanHaveBugs is a bit of a hack, no? [10:29] mwhudson, yes; created just so that we can have the IBugTarget views also available for things that don't directly provide that interface but can be adapted to it [10:29] salgado: what's the longer term strategy there? [10:30] mthaddon: hi [10:30] [.*(?:\\[(?:[pP]=[^ \\t]+,[ ]*)?(?:[rR][sS]?=[^ \\t]+)|(?:[Tt][Rr][Ii][Vv][Ii][Aa][Ll])\\])]' [10:30] mthaddon: what should I put in to make both you and PQM happy, here ;) [10:31] lifeless: https://code.edge.launchpad.net/~launchpad-pqm/lp-production-configs/trunk will give you ideas about ones that have worked in the past [10:31] lol [10:32] can we have pages registered for IBugTarget and have that work for things that can only be adapted to IBugTarget ? [10:32] AIUI, no. [10:32] However, remember that I know nothing. [10:34] salgado: bzr+ssh://bazaar.launchpad.net/~mwhudson/launchpad/project-to-bugs-adapter fixes the feeds tests btw [10:35] deryck: https://bugs.edge.launchpad.net/malone/+bug/606191 [10:35] <_mup_> Bug #606191: filing a bug created two bugs [10:35] spm, can we get any debug info on why loggerhead is crashing to mkanat? === al-maisan is now known as almaisan-away [11:32] bigjools, can I solicit some help from you? [11:33] rockstar: of course, how may I assist on this fine day in Prague? [11:33] bigjools, take a look at the diff here: https://code.edge.launchpad.net/~rockstar/launchpad/bug-602333/+merge/29679 [11:33] bigjools, basically, I need to make sure that distroseries that we create for source package recipe builds have a nominatedarchindep (so it doesn't use the arm builder) [11:35] rockstar: "ProcessorFamily.get(2)" ? [11:35] bigjools, however, if I try and set the processor or architecturetag on the makeDistroArchSeries, there's a constraint violated. [11:35] bigjools, I tried that, but it violates a unique constraint. [11:36] I'm assuming this means that there's possibly some sampledata somewhere that I can use instead. [11:36] quite probably [11:36] which constraint BTW? [11:38] bigjools, looking, one sec. [11:50] salgado: Total: 4451 tests, 40 failures, 12 errors in 59 minutes 50.391 seconds. [11:50] from ./bin/test -vvc lp.bugs [11:50] salgado: not too bad, i guess [12:07] mthaddon: its landed [12:07] mthaddon: what happens next? === Ursinha is now known as Ursinha-afk === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [12:47] lifeless: you request a rollout via the LPS wiki page - we still need to talk to flacoste about whether you can be added to the list of approvers for that [12:53] mthaddon: thats on wiki.canonical.com yeah? === Ursinha-afk is now known as Ursinha [12:59] mthaddon: yes, that's fine, lifeless has the responsibilities than Bjorn previously [13:00] flacoste: cool, thx === almaisan-away is now known as al-maisan === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [14:51] lifeless: how's prague treating you? [15:15] rockstar: https://bugs.edge.launchpad.net/launchpadlib/+bug/316694 [15:15] <_mup_> Bug #316694: Add web_link property to resources [15:37] mtaylor: could be better, could be worse :) [15:37] lifeless: well that's good I suppose [15:50] lifeless: so, I have a thing I just hit up against which may be a bzr thing or may be a launchpad thing [15:50] hi monty [15:50] lp:~leonardr/launchpad/true-anonymous-access - leonardr you legend :) [15:50] hi poolie [15:51] poolie: Laptop rescue successful? [15:51] yes, the water-cooling project was succesful :) [15:51] lifeless: I've got a branch protected by a PQM (lp:swift - using tarmac) [15:51] lifeless: so nothing hits trunk there without going through a merge proposal [15:53] thing is - if the only difference in the tree is that I just tagged a release [15:53] that's always been weird to push, but has sort of always "just worked" ... how do I do that in this workflow? [15:54] (I'm in here because the passive aggressive side of me is going to request a tag-the-branch feature be added to "make a release" ... or something) [16:11] mtaylor, basically you need to either connect in using a key that lets you get at the bot [16:11] or put it into a new branch that you merge [16:11] or perhaps tarmac should get a command to do this [16:11] leonardr, i think that's all the reviews i need to do for you? [16:14] poolie: so I can do a merge request where the only branch diff is a tag? cool [16:14] mtaylor, i think you need a new revision, but it doesn't have to have any file changes [16:15] poolie: ok. so I'd want to do "bzr commit --unchanged ; bzr tag blah" and then merge that [16:16] poolie, i need you to look at my revisions to https://code.edge.launchpad.net/~leonardr/launchpadlib/improve-workflow/+merge/29849 [16:16] right [16:16] _slight_ ugly in that that'll wind up with the tag in an internal commit [16:16] but will work [16:16] mtaylor: I don't know whether tarmac calls the right bzr apis to propogate tags or not. Check the code. [16:17] lifeless: I'm pretty sure its process is "bzr checkout trunk ; bzr merge ; bzr commit" [16:17] right see the bug filed last week about this [16:17] lifeless: oh good [16:17] bzr commit in a heavyweight checkout doesn't propogate tags. [16:17] it should [16:17] please fix! [16:17] rockstar: ^^^ [16:18] rockstar: although what I _really_ want is a way to tell tarmac to tag a version [16:18] mtaylor, can you file a bug about that? [16:18] rather than needing to merge in a tag revision [16:18] rockstar: yes [16:18] lifeless: we don't have tagging done in the hudson bzr plugin do we? [16:18] lifeless, it's using a lightweight checkout. [16:19] mtaylor, how would you expect to specify a tag? [16:19] * mtaylor goes to write a hudson param job that will tag a branch [16:34] rockstar: ok cool that should be fine then [16:48] rockstar: not sure [16:48] mtaylor, I think you probably should just tag after the fact. [16:48] rockstar: it's an element of trunk-protected-by-pqm workflow I've only just now thought about [16:49] rockstar: totally - the question is, if my hudson is the only one with permission to do it- what's the least ridiculous way to say "yeah, this one is going to be 1.0" [16:49] mtaylor, well, you always want someone to be able to get to the branch that is robotically managed. [16:49] what if there was a launchpad thing, similar to propose merge which was propose release... [16:49] rockstar: well, I can [16:49] For instance, our SAs can still edit the branch. [16:50] mtaylor, you should also make a bug about proposing a release. [16:50] rockstar: but in a general workflow/process description, "ssh in to the hudson machine as the hudson user, branch trunk, tag and push" is sort of silly [16:50] rockstar: I think I will [16:50] flacoste, that's great about "what's new", do you know there's a bug for that? [16:50] rockstar: I would be a nice workflow addition, potentially [16:50] Also, it might be cool if you could tag a branch through the LP web interface. [16:50] rockstar: yes. the github folks have already poked me about that [16:50] poolie: yes, it's bug 129943 [16:50] <_mup_> Bug #129943: Changing home page "What's new" item shouldn't require a code rollout or cherry pick [16:51] although I do not care for github's version of making a tarball for that :) [16:51] mtaylor, there's an open patch to loggerhead about making a tarball, but I was concerned that it might have memory issues. [16:52] (because loggerhead is REALLY good with memory, and we don't want to ruin that) [16:52] [16:52] rockstar: haha [16:52] mtaylor, truthfully, please file a bug on EVERYTHING the github folks poke you about. Worse thing that could happen is we could mark it "Won't fix" and actually provide a good explanation. [16:53] (As opposed to "No, that's stupid") [16:53] rockstar: I care less about making the tarball (as I'd want to use make distcheck or python setup.py sdist in any case) as I do about the workflow around deciding, taging and creating the relase object [16:53] rockstar: indeed === flacoste is now known as flacoste_lunch === al-maisan is now known as almaisan-away === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha === beuno is now known as beuno-lunch === beuno-lunch is now known as beuno [19:08] rockstar: tarmac is totally running the show now. complete rock === mordred_ is now known as mtaylor === flacoste_lunch is now known as flacoste [21:39] mtaylor, WOOT! [21:40] rockstar: I'm also getting a list of bugs to file (mostly feature requests) [21:42] mtaylor, nice. I'll see if I can get them fixed soon. [21:42] rockstar: the least silly one was someone asked if tarmac indicated anything in the commit about the reviewers/approvers [21:43] mtaylor, okay, so there's a way to construct a commit message template that can give that information, and the reviewers are also stored as revprops on the merge. [21:46] rockstar: oh they are? [21:46] rockstar: I think the revprops thing is probably fine [21:46] rockstar: I just did a bzr log and didn't see anything [21:47] mtaylor, yeah, so I need to write a bunch of documentation this weekend. [21:47] yay! [21:47] rockstar: the silly bug is that if you try to merge a branch that's already merged, it doesn't update the status to merged [21:48] rockstar: as in, somebody proposes for merge, there are no changes, it's approved, there are no changes - tarmac gets it ... it keeps sitting as approved in the merge queue :) [21:48] rockstar: amazingly enough, this did actually happen :) [21:48] mtaylor, it doesn't update the status to merged because the mp should be updated on launchpad to merge. [21:49] hey [21:49] rockstar: it should - but nothing ever gets committed locally [21:49] so nothing gets pushed and thus nothing gets triggered [21:49] hey thumper [21:50] mtaylor, okay. So the bug is "Pointless merge doesn't remove it from the queue" [21:50] rockstar: yup [21:50] rockstar: I'll file it and stuff... I just haven't gotten there yet [21:51] mtaylor, cool. [22:20] thumper, which room are you in? [22:20] 424 [22:21] thumper, coming down... [22:21] rockstar: in that case, 425 === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha