=== robru changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): cihelp | CI Train support: trainguards | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [02:04] === trainguard: IMAGE 149 building (started: 20140725 02:05) === [04:44] === trainguard: IMAGE 149 DONE (finished: 20140725 04:45) === [04:45] === changelog: http://people.canonical.com/~ogra/touch-image-stats/149.changes === [05:00] shame that the new UITK did not make it to dthe 149 image [06:38] hi gentle CI train operators, can I get a silo for line 24? [06:49] sil2100, good morning [06:57] sil2100, in case you're around ↑, bear in mind we've experienced a ~new Qt crasher with this silo, we'll be investigating it asap today, but we need it to land to get the code and dbg symbols [06:57] sil2100, did someone look at the test_no_sigstop failure? Saviq ? [06:57] brendand, it's fixed in ↑ [06:57] Saviq, silo009? [06:57] brendand, yes [06:57] Saviq, which is landing when? [06:58] brendand, I just marked it OK for landing [06:58] \o/ [06:58] gonna have a green build, yeah! [06:58] brendand, there's one more known failure there, but benign and isolated to unity8 [06:58] so not really [06:58] but we just can't block this silo any longer... [06:59] Saviq, ? [06:59] Saviq, so it fixes that test, but breaks which one? [06:59] brendand, http://ci.ubuntu.com/smokeng/utopic/touch/mako/149:20140725:20140717.1/9247/unity8/1417433/ [06:59] not the "fix" fixes it, but the UI changes in the silo [06:59] s/fixes/breaks/ [07:01] :( [07:01] brendand, we want to land this silo and deal with the issues after that, dealing with them from the silo is rather painful (lack of the actual output branch and dbg symbols) [07:02] Saviq, that's unfortunate, but ok [07:02] but we'll be tackling both today, and if we can't, we'll revert the offending landing [07:03] sil2100, good morning, can I get a silo for line 24 ? [07:05] sil2100, if Saviq can get unity8 green by EOD, we buy a beer for him every day between us at the next sprint :) ? [07:16] psivaa, can you try rerunning weather app? [07:28] can I get a silo for line 23? [07:31] ;) [07:31] thostr_, oSoMoN: give me a moment [07:32] brendand: the same test has failed before. There's likely something wrong while opening the toolbar. [07:33] brendand: so yeah, give me a moment to get up-to-date on the situation [07:34] elopio, ? too much coffee :) [07:35] brendand: no, I just want not to work tomorrow and take my holiday :) [07:35] well, it's already tomorrow, I failed a little. [07:42] elopio, is there a bug? has it been investigated/reproduced locally? [07:43] brendand: not that I'm aware of. My phone is flashing. [07:46] elopio, go to bed - i'll take care of it [07:49] brendand: I'm not yet done with the performance stuff, but I'm close. [07:50] brendand: I think we will need a screenshot to understand what happened there. But that's close to land on autopilot, I wouldn't worry about it to much. [08:05] elopio, if you're still here... I need guidance... if you remember the autopilot changes on https://code.launchpad.net/~saviq/unity8/drop-filtergrid/+merge/226415 [08:06] psivaa: hmmm, why smoketesting doesn't have results for 149? [08:06] psivaa: or wait [08:06] psivaa: nevermind! ;) [08:07] sil2100: I was about to call the ambulance :) [08:08] Saviq: I'm here. [08:08] let me see your latest changes. [08:09] Saviq: ok, anyway, I'll publish unity8 now - is that crasher user visible? [08:09] bah ... silo2 failed :((( [08:09] sil2100, yeah, it is, we're working on it now [08:09] sil2100, just don't want to block stuff any longer [08:10] elopio, sorry to be bothering you, but I'm lost over here... so what it does as hopefully you'll be able to read is comparing the globalRect.y of items that are outside of the category_element (which is collapsed) [08:10] and comparing that to the height of that category_element, determining which are actually in view [08:11] brendand: after some struggle with a device, weather app is running again [08:11] sil2100: I am pleased to see this : http://ci.ubuntu.com/smokeng/utopic/touch/mako/149:20140725:20140717.1/9247/ubuntuuitoolkit/ [08:11] bzoltan: yeah ;) It's excellent, really love our dashboard now [08:12] Just a bit sad about the Unity8 blocker still present... while I don't know if I want to introduce a user-visible crash [08:12] sil2100: what is wrong with the unity8 and the weather app tests? [08:12] sil2100: ahhh, so the unit8 problem is still there... shame [08:13] Saviq: you are not bothering. But it is becoming hard to read. I'll branch to get it running, that will be clearer. [08:14] bzoltan, weather app is being investigated, but it has been floating around for a while [08:14] bzoltan, it's not always failing [08:14] brendand: I see. [08:17] elopio, actually... I think I just understood the problem [08:19] elopio, I'll deal with it, thank you for being a sounding board [08:19] Saviq: unapproved merges in the unity8 landing! [08:20] Saviq: jeje, ok. I'll finish what I'm doing and if I'm still not sleepy I'll take another look at your branch. There are a couple of things I'm not sure I like. [08:20] Is this related to the crash? [08:21] Saviq: and unity8.shell.tests.test_emulators.DashAppsEmulatorTestCase.test_get_applications_should_return_correct_applications has just passed here, btw. I though you were having problems with that one. [08:21] elopio, yeah, I was [08:22] elopio, and has it passed on your phone? in silo 9? [08:22] elopio, or just that branch [08:22] Saviq: I haven't tried the phone. Just tried that branch on my desktop. [08:23] elopio, yeah, I know what's happening, thanks [08:23] sil2100, no, that one is just a tooling change [08:24] sil2100, it's approved now [08:26] ahhh, just one failure [08:26] brendand, good news, though, I know the failure and will have a branch in a moment [08:30] free beer for Saviq! [08:30] I just published unity8, but we want to get rid of the failure and the crash! FAST :D [08:30] sil2100, bug #1348531 btw [08:30] bug 1348531 in Ubuntu CI Services "Silo branches should be public" [Undecided,New] https://launchpad.net/bugs/1348531 [08:31] and one more incoming [08:43] tvoss: google+ is crashing a lot recently we are wondering if it might be related to the webbrowser location crash, I'm going to be digging into it today :) [08:44] bug #1348535 [08:44] bug 1348535 in Ubuntu CI Services "Need -dbgsym packages from silos" [Undecided,New] https://launchpad.net/bugs/1348535 [08:45] davmor2, the webbrowser location crash should be fixed [08:46] tvoss: not according to the latest test it isn't [08:47] davmor2, where do you see that crash? do we have a bug logged? [08:48] tvoss: not yet I'm going to try some more webapps like facebook and stuff [08:48] davmor2, sure [08:50] davmor2: http://ci.ubuntu.com/smokeng/utopic/touch/mako/149:20140725:20140717.1/9247/webbrowser_app/ <- here's the crash [08:50] tvoss: ^ [08:50] tvoss: ^ [08:51] davmor2, sil2100 so the browser did not crash, but the service? [09:01] davmor2, sil2100 if so, https://bugs.launchpad.net/location-service/+bug/1347887 [09:01] Launchpad bug 1347887 in location-service "Location service should install vanilla /etc/gps.conf" [High,New] === renato is now known as Guest59000 [09:03] sil2100, line 30, can I haz silo, please? [09:08] sigh, whats that ... in silo2 the installation of procps and udev when building the chroot fails ... [09:08] tvoss: hey! Ok, cjwatson is the planned sheriff, but let me do that anyway :) [09:08] ogra_: ugh? Let me see that [09:08] https://launchpadlibrarian.net/180716223/buildlog_ubuntu-utopic-armhf.ubuntu-system-settings_0.3%2B14.10.20140725.1-0ubuntu1_CHROOTWAIT.txt.gz [09:09] sil2100, thanks [09:10] Holy moly [09:11] sil2100, ? [09:11] tvoss: it was to the thing ogra_ pointed out, no worries ;) [09:11] * tvoss is scared of silo 8 :) [09:11] ogra_: I re-ran one build to see if it was transient [09:11] ogra_: ok, it doesn't seem to be... [09:12] It failed now as well [09:12] there was one arch where mountall was the issue ... all others had procps [09:17] sil2100, brendand, got fix for both issues [09:18] hi, sorry I'm a bit late this morning === cjwatson changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): cihelp | CI Train support: cjwatson | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [09:18] anything I should know from the meeting? I assume we're pushing towards another promotion ASAP ... [09:19] cjwatson, rigght, we do [09:19] there is one unity8 failure (and a few crashes) [09:19] unity8 was uploaded ... waiting for promotion to build a new image [09:20] waiting for migration to the release pocket you mean? [09:20] sil2100, I can't see unity8 in excuses, so IIUC it should migrate momentarily? [09:20] yeah [09:20] if you can't see it in excuses at all then it isn't being processed in the current run [09:21] Setting up sysv-rc (2.88dsf-41ubuntu16) ... [09:21] info: Reordering boot system, log to /var/lib/insserv/run-20140725T0903.log [09:21] success: Enabled dependency based boot system. [09:21] is that normal ? [09:21] oh, proposed-migration is stuck. will investigate [09:22] i thought we circumvented insserv [09:30] proposed-migration fixed [09:31] (my fault from changes last night, sorry about that) [09:32] I suspect everything is going to fail to build in utopic right now though [09:33] ooh? [09:33] apparently there's an apt bug that the base system just managed to tickle [09:33] so everything's falling over with "Chroot problem" [09:33] cjwatson, so it does (failing on setting up initscripts in my PPAs) [09:33] Adam's going to work on it shortly [09:33] k [09:33] But don't bother building silos in the meantime [09:34] greyback, mzanetti, bad news ↑ [09:34] well, i cant, they fail setting up the chroot :) [09:34] yep [09:34] Right, I mean don't bother trying :) [09:34] heh [09:35] well, its a good thing ... that gives us wiggle room for the new image so not to much lands [09:36] ogra_, but I *just* wanted to land a fix that would make it greeeeen! ;((((( [09:36] Saviq, oh ? i thought that was in the last upload [09:37] Saviq: is that unity8 from silo 9? [09:37] cjwatson, no, another one [09:37] cjwatson, ogra_, we wanted to land silo 9 and do a real quick apfix + crashfix landing [09:37] just after [09:38] ah, I see [09:38] well, hopefully it won't take too long to sort out [09:39] Saviq: :< [09:40] unity8 is migrating nowish [09:40] well, after p-m spends an age thinking about libav again [09:49] popey: Bill uploaded a gallery-app to the store some time ago that needs to be approved in order that we can unblock the libav transition, and it doesn't seem to be in place yet. Could you have a look please? Should be at least 2.9.1.1016. [09:49] popey, is out this week [09:49] (he pointed us to others that could do the job but i forgot who that was ... sil2100 do you remember ?) [09:50] cjwatson, ogra_: I know he pointed us to dpm and balloons [09:50] ah, dpm, right [09:50] With the exception that dpm, I think, doesn't have the power to generate click packages, but only approve them for the store or something [09:51] Ah, right, but Bill said he'd uploaded it so that shouldn't be necessary [09:51] dpm: ^- Could you have a look at that please? [09:52] cjwatson, sure, although I'll ask dholbach, as he's the one doing the reviews/approvals if popey is not there. Let me talk to him. [09:52] sil2100, cjwatson, it's about reviewing and approving gallery-app only, right? [09:59] cjwatson, sil2100, dholbach tells me that the latest gallery-app upload was already reviewed, but it was rejected because it uses the wrong framework. So someone needs to fix it and re-upload [09:59] Argh [09:59] Ouch [10:00] that was 1 day ago, with gallery-app 2.9.1.1018 [10:00] Wasn't it built by Jenkins? Does that mean Jenkins is using the wrong framework? [10:00] Or was it built manually? [10:00] no idea, oSoMoN, do you know how the gallery-app click is generally built? ^^ [10:02] tvoss, dbarth: okay so I've been opening a few webapps g+ is the only one crashing but in the log files on g+ and facebook I see a lot of libust Error: Error opening shm /lttng-ust-wait-5 (in get_wait_shm() at lttng-ust-comm.c:886) is this anything to do with location? Or is it an issue in webapps? [10:02] cjwatson, yes, it seems that trunk still has the deprecated framework: http://bazaar.launchpad.net/~phablet-team/gallery-app/trunk/view/head:/click/manifest.json.in [10:02] davmor2, ignore that one [10:03] ogra_: meh I'll keep digging then [10:07] dbarth, tvoss, ogra_: Sometimes I hate the phone http://paste.ubuntu.com/7854399/ [10:07] davmor2, :) [10:07] davmor2, apport-retrace works great on your host, too [10:07] lovely [10:08] tvoss: but for retrace to work you need to of viewed the file on the device first right? [10:10] davmor2, nope, just grab the .crash file, install apport-retrace, bzr branch lp:~ubuntu-archive/apport/lp-retracer-config, run sudo apport-retrace -o out.trace -S lp-retracer-config/ -C ~/retrace path_to_crash_file [10:10] davmor2, that gives you a meaningful t a a bt full in out.trace [10:10] davmor2, takes some time, though :) [10:22] tvoss: okay what am I missing? If I try a lowercase c or remove the -C altogether I get ERROR: report file does not contain one of the required fields: CoreDump DistroRelease Package ExecutablePath with the -C in place I get http://paste.ubuntu.com/7854452/ [10:23] dpm: Are core apps generally using 14.10-blah-dev3 now? [10:23] davmor2, -C ~/retracer, -C expects the cache Dir [10:23] davmor2, got the link to the .crash file? [10:23] tvoss: thanks [10:23] cjwatson, I think the community core apps are using 14.10-dev2 [10:24] we updated them all at once a few weeks ago [10:24] tvoss: http://people.canonical.com/~davmor2/_usr_bin_webapp-container.32011.crash [10:25] yes, ubuntu-sdk-14.10-dev2 [10:26] davmor2, okay, that indeed lacks a core dump [10:27] davmor2, that crash file indeed lacks a core dump [10:27] tvoss: so no use then :( [10:27] davmor2, nope [10:27] sil2100, u8 fixes are in line 32, so whenever silos are back, please expedite it [10:28] tvoss: see told you sometimes it's good to hate the phone [10:29] dpm, sorry, no, I don’t know, bfiller would know [10:29] it's like therapy [10:29] davmor2, ah okay, webapp container is a script, isn't it? [10:30] hmmm, it's not [10:30] ev, you around? [10:30] Saviq: o/ [10:30] Saviq: thanks! [10:35] sil2100, i gave up reproducing the weather-app failure [10:35] sil2100, for a little while it was really reproducible. now it's gone [10:35] brendand: maybe it was just something singular [10:35] Oh [10:35] It was? [10:35] sil2100, yeah [10:36] sil2100, next time we see it i'll file a bug [10:36] sil2100, are we pushing for our green image today? what's the outlook? [10:37] brendand, broken chroots everywhere ... we're waiting [10:38] ogra_: chroot issue is fixed. I'm retrying a mess of builds. [10:38] awesome ! [10:38] brendand: yeah, the chroot problem was standing in our way [10:38] infinity: oh, can we press rebuild now ;) ? [10:38] i thought it was waiting for another publisher run [10:38] sil2100: If by "press rebuild" you mean "retry the builds in LP", sure. If you mean "pointeless reupload", please don't. Cause I'm doing the former. [10:39] infinity: no no, just retry rebuild in LP [10:39] infinity, silo (PPA) rebuilds ... [10:39] * ogra_ tries silo 2 [10:39] ogra_: Too often when someone says "rebuild" here, they mean "re-do the whole thing". [10:39] Cause there's a button to press that makes that easy. [10:40] not in here ;) [10:40] sil2100, will unity8 pass now or does it need more landings? [10:40] brendand: it needs one more landing - now things should be better, but we need one more landing to fix a crash [10:41] dpm: or ubuntu-sdk-14.10-qml-dev2 ? [10:41] Most everything should be retrying on all arches by now. I'm just cleaning up x86. [10:41] mind you, I guess gallery-app isn't really just using the qml framework [10:42] sil2100, and is that silo created, or just in MP form at the moment? [10:42] brendand: it's in a silo :) [10:43] Saviq: ^ [10:43] * infinity looks at that suspiciously. [10:43] Saviq: no commit message for one of the merges! [10:43] sil2100, that's silo009 again i guess [10:44] sil2100: I think we can kick an image after this publisher run (has unity8 and url-dispatcher) - shall I take care of that, or do you want to wait for silo 9? [10:44] cjwatson: I would love silo 9 before that... I hope we can have it landing in the next 20-30 minutes [10:44] ok [10:45] Saviq: ! [10:45] Ok, I set a quick commit message [10:46] sil2100, oops, thanks [10:48] sil2100, i'm at hand to do anything needed to expedite things - testing, checking, rechecking etc [10:48] brendand: excellent! Saviq might be happy for some help once silo 009 finishes building [10:49] oSoMoN: https://code.launchpad.net/~cjwatson/gallery-app/framework-14.10-dev2/+merge/228273 [10:49] ARGH, FFS. [10:49] sil2100: Who did the new upload of ubuntu-system-settings just now? [10:49] sil2100: The exact thing I asked people not to do. [10:50] ogra_: ^-- That's what I was talking about. :( [10:50] infinity, err, that was me ... and i didnt do an upload [10:50] cjwatson, I believe most, if not all of them use ubuntu-sdk-14.10-dev2 [10:50] ogra_: You pushed a button that generated an upload when I just told you the build were retrying. [10:50] Pressing the "Build" button in CI Train is doing an upload. [10:50] sorry [10:50] The fact that a robot is doing it for you doesn't change that. [10:51] This is why we can't have nice things. [10:51] cjwatson, actually, either 'ubuntu-sdk-14.10-qml-dev2' or 'ubuntu-sdk-14.10-dev2' [10:51] * sil2100 didn't upload anything besides building a completely new package [10:51] dpm: I made it the latter since -qml- is kind of a lie for gallery, it uses more than just that [10:51] ok [10:53] infinity: guess we could cancel the earlier-version builds [10:53] oh, too late [10:53] sil2100, so we can build silos again/ [10:53] ? [10:54] cjwatson: Not too late to save arm64 some time, doing. [10:54] cjwatson, approved [10:54] Saviq: yes, but please do not 'rebuild' any failed builds [10:54] Saviq: those have been automatically rebuilt in PPAs by infinity [10:54] sil2100, yup ok, please recon silo 15 though [10:54] we'll need to do watch-only builds in some cases, but I'll sort that out once builders have caught up [10:55] oSoMoN: yay, thanks. I can't really test this myself - any chance you could organise a landing? [10:55] oSoMoN: in the past bfiller has wanted things merged to trunk before building a click package for upload to the store [10:56] cjwatson, that’ll be too tight as I’m leaving for the week-end in an hour from now, would you mind pinging bfiller when he comes online, he can take care of the landing himself [10:57] * infinity goes to eat his whatever-meal-this-is. [10:57] oSoMoN: ok [10:57] thanks for the help [10:57] yw [10:57] cjwatson: queues for arm64 and ppc aren't remotely pretty, if people have urgent things, you may need to go on a rescoring romp. [10:58] yeah [10:58] I'll fix PPC next week, and start costing what it'll take to fix arm64. [10:58] \o/ [10:59] ^- that's just a watch-only build [11:01] davmor2: the lltng messages are an annoyance, but they are not cause for errors [11:02] davmor2: did you manage to post the crash file somewhere to be retraced? [11:02] dbarth: no coredump [11:03] dbarth: really frustrating though, g+ just won't stay open [11:04] (6 is also watch-only) [11:04] let me try it from the browser instead [11:06] ok, that's the last couple of watch-only builds kicked off to clean up after the chroot breakage [11:07] Saviq: huh, the previous build of 6 wasn't enough then? [11:08] cjwatson, we need to build the -gles packages after the real ones go in [11:09] Saviq: ah, gotcha [11:09] looks like it hates you though [11:10] cjwatson, yeah, rightfully so [11:10] fixed by now [11:10] greyback, did you sync qtubuntu-gles, too, shall I build both? [11:10] can I get a silo for line 34? [11:10] ah too quick ;) [11:10] Saviq: yes but I kicked off a build just for it, since you had qtmir going separately [11:11] thostr_: hang on, you're in the queue [11:11] Saviq: oh qtmir-gles filed [11:11] greyback, yeah, but I broke it, will do qtmir-gles alone after it uploads [11:11] cjwatson: ack [11:11] greyback, 's fine [11:11] mandel: do you want line 12 or line 31 first? (alternatively, which will you be able to land more quickly? we're low on silos) [11:11] tvoss, dbarth: might of found a work around I crashed the webbrowser on googleplus instead :) and that is uploading to a bug report hopefully \o/ [11:12] cjwatson, let me check [11:12] * sil2100 goes for lunch [11:13] cjwatson, 12 first since it adds mms [11:13] cjwatson, 31 later [11:13] mandel: ok [11:15] cjwatson, AFAIK 12 was tested and should have a +1, let me check with sergio [11:15] thostr_: I want to keep one free for emergencies, but if you can test 4 then we can get it published which will leave you next in the queue [11:15] mandel: well, I hope so, I just assigned it :) [11:15] cjwatson: ok [11:15] barry: any progress on testing 20? [11:17] silo 16 landed at the start of the week [11:17] wut [11:17] spreadsheet must be corrupt [11:17] damnit [11:18] tvoss, dbarth, sil2100: one bug for google+ dying, https://bugs.launchpad.net/ubuntu/+source/webbrowser-app/+bug/1348592 [11:18] Error: launchpad bug 1348592 not found [11:18] logged in right on time to get the hilight; I usually don't check if I don't know about something :-) [11:18] made it public now so you should be able to get at it :) [11:18] ok, evicting 16 [11:19] cjwatson, it was tested.. I mixed line number with silo number (stupid me) [11:19] davmor2: I think that's a dup on an oxide bug [11:19] sil2100: could you sort out the spreadsheet state for line 12 when you get a minute? it should be landed [11:19] mandel: so 31 is actually still outstanding? [11:19] sergiusens: could be but I thought I would file it separate to be sure, there were a bunch of bugs that looked similar [11:20] cjwatson, 31 needs testing once 12 has been published [11:20] cjwatson, although the changes are super safe, it just updates the logging [11:20] mandel: 12 was published some time ago [11:20] mandel: as sergiusens says [11:21] cjwatson, sweet, then I'll wait a little and will test 31 [11:21] mandel: ok, I'll assign a silo for that shortly, you can't test it yet as you have to build it first :) [11:21] cjwatson: line 12? Let me take a look [11:21] cjwatson, yes, I need the silo to follow the correct path :) [11:22] cjwatson: hm, but it's being freed right now, right? So the changes haven't been merged in? Or did they? [11:22] sil2100: it's cleaning now from silo 16, but it should've been flagged as landed way before that - it actually landed on Tuesday [11:22] sil2100: so I was misled into assigning a silo for it [11:22] Oh [11:22] sil2100: didn't actually build anything in this run though [11:22] Ok, sure, let me try fixing that === tvoss is now known as tvoss|test === tvoss|test is now known as tvoss [11:26] tvoss: what are you doing there? [11:27] cjwatson, I have hit rebuild ... sorry for that [11:27] tvoss: I already sorted it out after the chroot glitch [11:27] cjwatson, oh sorry [11:28] tvoss: so let me do a second watch-only build now to get it back into the right state, unless you actually have new source changes to build? [11:28] cjwatson, nope, no actual changes [11:28] Saviq, brendand: could you guys give silo 9 a spin? It's ready [11:29] tvoss: right, done [11:29] cjwatson, thanks [11:29] mandel: building silo 16 for you [11:30] cjohnston, \o/ [11:31] sil2100, doing [11:32] alexabreu: Any luck with testing silo 11? It's been there for a while and we're low on silos [11:36] sil2100, +1 [11:37] yay, publishing [11:39] cjwatson, can you make out anything out of https://ci-train.ubuntu.com/job/landing-006-1-build/149/console ? [11:40] Saviq: that's the same problem that bit the archive proper earlier; needs a chroot update in jenkins to work around it [11:40] ! [11:40] :) [11:40] Saviq: perhaps fginther could coordinate with infinity to get that fixed [11:40] \o/ [11:40] \o\ [11:40] /o/ [11:40] (it's an apt bug, but avoidable with careful chroot updates) [11:41] cjwatson, should I just try again? [11:41] Saviq: won't help in the slightest [11:41] ugh :| [11:41] cihelp: ^- if anyone's available [11:44] bfiller: Can you make anything out of the pile of failures in https://code.launchpad.net/~cjwatson/gallery-app/framework-14.10-dev2/+merge/228273/comments/552603 ? [11:44] davmor2, some non-uploader testing of silo2 would be appreciated [11:57] ogra_: will do after Lunch [11:57] thanks !!!! [11:58] * ogra_ hugs davmor2 === alan_g is now known as alan_g|lunch [12:06] cjwatson: hey, can I get a silo for l35? [12:06] ah, remaining silos 1 [12:07] well, thats enouogh for one line, isnt it ? [12:07] :) [12:07] sergiusens: well, there's a queue [12:07] and I was hoping to keep one free for emergencies [12:08] cjwatson: yeah, my "ah. remaining silos 1" was supposed to be followed with a :-( [12:08] I'll do thostr_'s once unity8 finishes landing [12:16] sil2100, sorry, was out for a bit [12:16] sil2100, i can do that if still needed [12:16] brendand: no need, it's published! [12:16] sil2100, :) [12:16] sil2100, and we kicked a build? [12:17] Not yet, didn't migrate completely :) [12:17] But once it does, we kick a new image and do promotion dogfooding with it! [12:20] publishing that as it's not on images [12:20] cjwatson, thanks [12:24] unity8 waiting for autopkgtests now [12:24] we really need to finish landing libav so that proposed-migration can be fast again :-/ [12:25] cjwatson, what changed in libav? [12:25] out of curiosity [12:25] 9 -> 10 transition [12:25] New version, they changed the ABI [12:25] but it's ended up intertwined with a bunch of other library changes in the archive [12:26] which now all have to land at once [12:27] we have to land it before we branch for RTM or it'll be a nightmare [12:27] fortunately it really is close now [12:28] cjwatson, okay [12:28] cjwatson, sounds complicated :) [12:29] 174 source packages at the moment [12:30] oh god and something's changed again so it generates 10418 new uninstallables. Hopefully it's just transient === tvoss is now known as tvoss|lunch [12:36] ah, was just waiting for curl/armhf build, good [12:43] cjwatson, for silo 11 I'll prob close it today === psivaa is now known as psivaa-afk [12:43] Is there a high level document/wiki explaining how Jenkins works to a newbie? [12:43] cjwatson, I just asked silo-18 to build and it seems nothing happened [12:45] pmcgowan: it's already built? [12:45] cjwatson, the branch changed so trying to rebuild [12:46] pmcgowan: the first time you click build on the jenkins job it often ends up having to log you in, and you should then press the button again [12:46] yay jenkins [12:46] ok will try again [12:46] alexabreu: thanks [12:46] cjwatson, I seem to be missing the build permission [12:47] pmcgowan: yeah, you aren't the lander on that [12:47] sil2100, can you please reconf silo 15, it's got a messaging app added [12:47] pmcgowan: building it for you now [12:47] cjwatson, ty sir [12:47] cjwatson: could you? ^ I'm hacking poor jenkins now :) [12:48] sure [12:51] Saviq: done [12:52] cjwatson, thanks [12:56] ogra_: anything in particular I'm looking at with silo 002 or just testing in general? [12:57] davmor2, general testing and it would be good to make sure that toggling the developer mode switch in "About->Developer mode" actually switches adb on and off (mtp reconnects are expected) [12:58] * davmor2 lends sil2100 a big axe to help hack jenkins [12:58] ogra_: will do [12:58] thanks :) === alan_g|lunch is now known as alan_g [13:00] oh I like the new see all much better than the arrow that always seemed to point the wrong way :) [13:00] oh buttons are suddenly blue instead of orange is that meant to be the case? [13:00] * ogra_ doesnt know [13:01] (my flo is still flashing) [13:03] davmor2: this axe will be helpful indeed! [13:03] Especially that... the firewall is blocking access to the dogfood instance :| [13:04] ogra_: okay I'm confused! I read the following "You need a passcode or passphrase set to use developer mode", then in lock security I have the option for Swipe (no security) Surely that should be greyed out while in developer mode [13:04] davmor2, yes, but we cant do that yet [13:04] meh [13:04] sil2100: sorry to pester, i tried to read scrollback, but.. silos aren't building ? [13:04] sil2100: ah, #webops should hopefully be able to help [13:04] kgunn: which ones? [13:05] davmor2, this landing is solely the UI changes and the switch connected to the usb driver [13:05] cjwatson: i'm just getting on for mornin', so if you say "go build" i will [13:05] kgunn: which ones? [13:05] cjwatson: silo6 [13:05] morning or not I need details :) [13:05] ogra_: is the lock meant to work? [13:05] kgunn: ok, that needs a CI vanguard to wake up [13:05] davmor2, then i'm waiting for silo12 ... once that landed the lock settings will be made active [13:05] cjwatson: also silo12 [13:05] davmor2, only after silo12 landed [13:05] kgunn: same [13:06] 12:40 Saviq: that's the same problem that bit the archive proper earlier; needs a chroot update in jenkins to work around it [13:06] 12:40 Saviq: perhaps fginther could coordinate with infinity to get that fixed [13:06] 12:40 (it's an apt bug, but avoidable with careful chroot updates) [13:06] cjwatson: thanks... infinity ^ is this possible ? [13:06] ogra_: so if I citrain silo 12 aswell it should all be there and work right? [13:06] oh wait, i see francis is on [13:07] davmor2, no [13:07] davmor2, there is no code for this yet [13:07] davmor2, only toggling adb on/off should work atm [13:07] kgunn: infinity can't do it directly, I only highlighted him because he can probably offer advice to whoever's fixing up the pbuilder chroots [13:07] ogra_: sadtrombone.com [13:07] kgunn: please don't bug him [13:07] davmor2, the rest will land early next week [13:07] ogra_: right no worries [13:09] oh man [13:09] olli: ^ most likely looks we won't be landing qtcomp, locking greeter is a maybe [13:09] i see what you mean with the colors [13:09] looks like candyUI now [13:10] cjwatson: i've tested it. dunno if others have yet (morning email, here i come) [13:10] barry: who else does it need? [13:10] ogra_: dev mode off, mtp works, restart, mtp works, and so does adb [13:11] and adb is also off with dev-mode off i hope :) [13:11] ogra_: and I go into developer mode and it is turned on again [13:11] cool [13:11] that is how it should be [13:11] cjwatson: nobody in particular. i'd just like confirmation that the test plan works for someone else too [13:12] ogra_: so dev mode should be back on after a reboot? [13:12] davmor2, yeah, thats hardcoded in the android setup still ... next step is to turn it off [13:12] (i need to talk to sergiusens about that) [13:12] ogra_: right okay well when I turned it off I got the following [13:12] root@ubuntu-phablet:~# davmor2@boromir:~$ adb shell [13:12] error: device not found [13:13] and mtp restarted [13:13] thostr_: assigning for you now, but it's possible that the build will fail for similar reasons to the above [13:13] yes, as i said, thats expected [13:13] looks fine [13:13] cjwatson: I'll give it a try [13:14] barry: ok, perhaps you can ask around for assistance, QA if you need it I guess [13:14] ogra_: yeap I'm saying it is good, So even though I was connected via adb that connection was dropped when I turned off developer mode which is a good thing [13:14] davmor2, dev mode really turns off the hardware for adb ... (which requires that mtp on the same wire gets restarted too) [13:15] davmor2, thanks ... setting to testing done then ;) (unless my own tests find anything) [13:15] ogra_: who said I'd finished :P [13:15] heh [13:16] btw, if you call phablet-shell with dev-mode off it should actually wait til you switch it on again [13:17] ogra_: you can quickly fill up your screen with nautilus windows by hitting that switch :) [13:17] davmor2, heh, yeah, blame nautilus/gvfs [13:17] i expect this is a swithc that gets only used rarely though ... [13:18] (unless you actually test the switch indeed) [13:19] ogra_, thx for fixing the deps in your branch [13:19] kenvandine, well, sorry for breaking them in the first place [13:19] :P [13:19] ogra_: I was hitting it repeatedly to see if it broke, it didn't, but it did cause error dialogs on the on the host machine :D GVFS isn't very adept at handling a device appearing and disappearing before it has connected it seems :D [13:20] yeah, it is pretty awful ... [13:20] (my testing was fine here too) [13:21] ogra_, great, you tested call forwarding too? [13:21] ogra_: EEEWWWWW slide down the battery indicator while it is on charge you get a lovely combo of the blue and green [13:21] kenvandine, only got a flo here [13:21] ok, i'll test that [13:21] davmor2, lol [13:21] please flip back to testing: no then until that's done [13:21] oh my [13:21] cjwatson, indeed [13:22] ogra_: okay I've tried the corner cases I can think of and everything is looking good from what I can tell [13:24] hmm, with the new battery indicator we dont have any quick way of setting the brightness anymore :( === tvoss|lunch is now known as tvoss [13:26] kenvandine, heyo! I want to land that locking branch of system-settings today. Any conflicts on your side? [13:26] we have settings in a silo about to be published [13:27] should be a nice flow today for system-settings ;) [13:27] and we have several other branches i'd like to get landed after this is published [13:27] mterry, can we squeeze your's in the middle there? [13:28] kenvandine, sure? [13:32] * ogra_ hugs kenvandine [13:33] sil2100, will we get a build in time for the landing meeting? [13:33] brendand: I don't know... unity8 is taking some time to migrate [13:34] it's migrating at the moment, sort of [13:34] should be in time for the landing meeting [13:34] brendand: one autopkgtest still needed to finish [13:34] maybe only just [13:34] sil2100: (I'm looking at the run in progress, which you can't see yet :-) ) [13:34] Hah ;) [13:35] cheating !! [13:35] ogra_, oh... wait your last change moving the new dep from build dep to depends wasn't built... [13:35] (well, you actually can, you have to look under log/utopic/ [13:35] ) [13:35] * kenvandine rebuilds [13:35] kenvandine, gar [13:35] * kenvandine hugs the packaging ack step :) [13:39] hm, it might depend on publisher runtime [13:39] it's really hard to tell just now, we're a bit swamped by KDE changes [13:40] ah, there, it will migrate next publisher run [13:44] kenvandine, jgdx branch is fixed I hear, should we add to silo 2 or do that next [13:46] pmcgowan, not yet... silo 2 is ready to publish after a rebuild [13:46] i'll get it in another silo after that [13:46] along with a few other branches [13:46] we already tested call forwarding and developer mode in that silo, just need a rebuild for a packaging change [13:47] * sil2100 spams rmadison [13:48] kenvandine, awesome [13:48] sil2100, silo009 migrated [13:48] kenvandine, btw did you run the wizard? [13:49] brendand: still needs to publish though [13:49] sil2100: at least spamming rmadison is no longer harmful :) [13:49] pmcgowan, not in this silo... [13:49] (well, apart from a little CPU time) [13:50] why is i always think rmadison is a person? [13:50] didn't remove any plugins or change the wizard... but now i guess i should :) [13:50] brendand: it was originally a person's name due to a terrible terrible habit of naming archive tools after women [13:51] cjwatson, for reasons of workplace sensitivity, it's probably best you don't elaborate on why it was a terrible idea :) [13:52] * cjwatson disclaims responsibility [13:52] * brendand too === josepht changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): josepht | CI Train support: cjwatson | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [13:52] but originally "madison foo" was what you ran in a shell on ftp.debian.org to find out which suite a package was in [13:52] or suites [13:53] and then madison-lite was my version of that that worked it out from Packages and Sources files rather than requiring direct DB access, after direct DB access was taken away from most Debian developers [13:53] and then somebody wrote rmadison as a client for a CGI script so you didn't need either a shell somewhere or to download the index files [13:54] sil2100: (before http://anonscm.debian.org/cgit/users/cjwatson/madison-lite/commit/?id=0031c4c97f420779338feb8c96ff11921b2f80c3, it was possible for spamming rmadison to increase the probability of stale cache files) [13:54] pmcgowan, ok, wizard was good :) [13:55] kenvandine, ;) [13:55] cjwatson: oh, I didn't know about that! [13:55] cjwatson, i guess it's safe to say that one of the reasons it's terrible is because sentences like 'x spams rmadison' are so easily misinterpreted [13:55] and go no further [13:55] cjwatson: damn, good thing that we didn't use rmadison in any scripts that would run it every 5 seconds [13:55] sil2100: nor did I until I worked it out :-) [13:55] well, I sort of knew something was weird sometimes, but it took a while [13:56] sil2100, didn't you notice it was horribly slow? [13:56] brendand: runtime was quite separate [13:56] that got a lot better with http://anonscm.debian.org/cgit/users/cjwatson/madison-lite/commit/?id=b2dad70fce28a85ba6ee1506a5b9b92a0e56de81 [14:03] Ok, let me kick a new image now [14:03] rmadison still says it's in -proposed, but I guess it should be safe now anyway [14:03] sil2100, i'm giddy with anticipation [14:04] sil2100, ooh, are you sure? [14:04] ...not completely, that's why I still didn't press the button ;) [14:04] Let me give it a few more minutes [14:11] sil2100: You weren't fine when you said that, but you are now [14:11] rmadison will be lagging a bit at the moment basically as an annoying side-effect of proposed-migration being slothful [14:13] sil2100, let's go! [14:14] theres an awful lot of preparing packages and not much building [14:14] pmcgowan: yes see above [14:14] needs a CI vanguard to wake up [14:14] ah, we have one now [14:15] josepht: we need to get the pbuilder chroots used by citrain upgraded to get past an apt bug. can you help with that? [14:16] cjwatson: I'm happy to help but I don't think I have permissions to do that [14:16] josepht: can you find me somebody who does? [14:16] cjwatson: sure [14:17] fginther: do you know who can handle this request? ^ [14:17] Building! [14:18] davmor2: please allocate some time for dogfooding the image that's building right now [14:19] josepht, I might be able to help... [14:19] catching up first [14:20] infinity: Could you help out fginther with what needs to be done to the pbuilder chroots? Should roughly match what you did this morning, although I know the archive has moved on. [14:20] cjwatson, do the chroots need to be recreated from scratch, or is it sufficient to just to an upgrade? [14:21] fginther: It'll be possible to upgrade them, I believe, but needs some manual care. Adam should be able to advise [14:21] fginther: Just upgrading is fine. [14:21] cjohnston, infinity, thanks [14:21] fginther: If they don't have proposed enabled, they should Just Work. [14:21] infinity: Oh, won't that run into the same problem? [14:21] err cjwatson thanks [14:21] any time [14:21] infinity: Since init was migrated [14:21] cjwatson: Did it migrate? Ahh, then it'll break, but a second run after an 'apt-get -f install' will work. [14:22] sil2100: Oh I don't know about that ;) [14:22] fginther: apt-get update ; apt-get dist-upgrade ; apt-get -f install ; apt-get dist-upgrade [14:22] Aha, good, thanks [14:22] davmor2: grrr! [14:22] :) [14:22] :) [14:22] infinity, got it [14:24] === trainguard: IMAGE 150 building (started: 20140725 14:25) === [14:27] cjwatson: Since it's migrated, I should probably refresh the chroots again, or the same upgrade bug might pop up the next time someone touches sysvinit, udev, util-linux, or mountall. But I'll do that after I sleep. And maybe mvo will have figured out the bug by then. [14:27] infinity: *nod*, thanks === alan_g is now known as alan_g|ea === alan_g|ea is now known as alan_g|tea [14:35] cjwatson, can i get silo for #36? [14:36] mhr3: you are in a queue, but I'll try to move the one ahead of you on now === boiko_ is now known as boiko [14:45] fginther: Let me know when that's done and we can try running something through ... [14:46] cjwatson, I'm doing a dry run to make sure I don't make it worse [14:46] OK === alan_g|tea is now known as alan_g [14:55] cjwatson, dry run is ok, doing it for realz this time... [14:56] cjwatson, this is only needed for the utopic chroot, correct? [14:57] cjwatson, it's done now. please give it a test [15:05] kenvandine, let me know when your silo lands so I know when to merge trunk into my branch again [15:06] mterry, will do [15:06] just waiting for the chroots to be fixed [15:06] kenvandine, right... :( === psivaa-afk is now known as psivaa [15:18] what's going on with the chroots? [15:18] broken [15:18] train wreck [15:18] ugh [15:18] derailed [15:18] offroad-chroots [15:25] ;) [15:27] sil2100, robru, are the chroots still broken? [15:27] From what I heard things should be good now [15:27] But not sure if that didn't change [15:28] sil2100, I applied the update as suggested about 30 minutes ago [15:28] kenvandine, ^^^ try a new rebuild ;) [15:29] sil2100: i'm selling an idea, can we subtract 1 day off the traincon-0 counter ? [15:29] since the chroot thing sort of torpedoed alot of planned landings [15:29] kgunn, so go into traincon-0 right now then ? [15:29] kgunn: well, theoretically we might have a promotable image today [15:29] ogra_: sorry, bad math...add a day back into the counter :) [15:29] ;) [15:30] kgunn: let's think about it once we get some feedback on #150 [15:30] sorry for being so german :) [15:30] hah ;) [15:30] ogra_: no no...sorry for being so american :P [15:30] lol [15:30] ogra_: i coulda been a doctor if it wasn't for all that science and stuff :) [15:31] hahaha [15:31] * kgunn also speaks 1 language [15:31] ogra_, will do [15:32] kenvandine, can you kick 18 as well [15:32] sure [15:32] sil2100, don't we already? what's wrong with 149? [15:33] brendand: #149 had the unity8 bug that caused it not to run all autopilot tests, which was marked as a blocker [15:33] brendand: so basically we never had a full unity8 test suite run [15:38] sil2100, yeah i guess so. well if 150 is the magic we think it is, then we'll be safe [15:38] sil2100, but what if it's not? will we really go to traincon-0? [15:40] brendand: yeah, but seeing the current image state, even if we go traincon it will most probably be only for up to a day at max [15:40] So not a big problem [15:43] fginther: thanks! Sorry, I was out for a bit [15:44] * cjwatson starts retrying failures [15:44] fginther: and yes, just utopic [15:45] * sil2100 looks at imgbot with anger, as if it would help speeding up the image build [15:45] imgbot, stunt [15:45] * imgbot rolls on its back and purrs [15:46] I guess I can ignore that for now as it was ignored in the last (failed) build pass === gatox is now known as gatox_lunch [15:46] * ogra_ guesses rsalveti is interested in that one [15:48] ogra_: you need a stunt-twang that leaps out at sil2100 twangs his sulky bottom lip and runs away [15:48] sil2100: any reason I shouldn't publish 4 and 13? [15:48] lol [15:48] cjwatson: no, I guess it's good to publish those if they're ready [15:49] apparently so, ok [15:49] ;) [15:49] dbarth: please top-approve merges you want published (silo 13) [15:51] sil2100: robru ....do you need to do anything special wrt turning silo6 into a landing silo ? we're keeping an eye on other silos and rebuilding... [15:52] bfiller: Are you around? Would you mind organising a landing for https://code.launchpad.net/~cjwatson/gallery-app/framework-14.10-dev2/+merge/228273 ? [15:52] bfiller: (Assuming you understand what those Jenkins failures are ...) [15:52] kgunn, no, there's no such thing as a 'landing silo'. they are all landing silos. you guys just hog silos and pretend they aren't for landings a lot. === josepht changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): josepht | CI Train support: cjwatson | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [15:52] Which BTW would be much better done using separate (non-silo) devirtualised PPAs. === josepht changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): cihelp | CI Train support: cjwatson | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [15:53] bfiller: (This is to fix the reason gallery-app was apparently rejected from the store, so it's still blocking the libav etc. transition) [15:53] kgunn, that said, just let us know, and we'll stop automatically ignoring conflicts with silo 6. [15:53] cjwatson: not sure what those jenkins failures are (all the AP tests are failing) but I'll investigate [15:54] robru: thanks ... "you guys just hog silos and pretend they aren't for landings a lot" ;) [15:54] cjwatson: will run the click on the devcie and see if the AP tests are passing [15:54] bfiller: thanks a lot [15:54] cjwatson: np [15:54] kgunn, that was perhaps less polite than I intended. [15:54] sorry [15:55] bfiller: (the click package ought to be built on utopic then, of course) [15:55] IMPOLITE robru! [15:55] sil2100, slag off! [15:55] ;D [15:56] cjwatson: yes, in fact it's built already by the jenkins job at this link http://jenkins.qa.ubuntu.com/job/generic-click-builder-utopic-armhf/320 [15:56] robru: no worries...just typical engineers, bending it to our will [15:56] bfiller: right, cool [15:56] hmmmmm [15:57] sil2100: it's in the system-image stage now FWIW [15:58] cjwatson: will we need to go through same exercise as last time, silo->proposed->merge to trunk->release click from trunk? [15:58] bfiller: Yeah, if you want the click package to be built from trunk; I assume the rationale there isn't going to have changed [15:58] And it made sense [15:58] cjwatson: ok [16:00] sil2100: need a silo for line 33 please [16:02] * sil2100 zaps davmor2 [16:02] * davmor2 hands the dogfooding to sil2100 now it's all on his head [16:04] * ogra_ hands sil2100 some shampoo ... to get the ground meat out of his hair [16:04] === trainguard: IMAGE 150 DONE (finished: 20140725 16:05) === [16:04] === changelog: http://people.canonical.com/~ogra/touch-image-stats/150.changes === [16:04] \o/ [16:05] davmor2: dogfood 150! [16:05] PLEASE [16:06] sil2100: ow my eyes the blue progress bar [16:06] heh [16:06] candy UI ... [16:15] _usr_lib_arm-linux-gnueabihf_url-dispatcher_update-directory.32011.crash [16:19] rsalveti, hmm, image 150 doesnt boot anymore [16:19] ogra_: yeah, will check [16:19] ah, just saw that sil2100 was evil and poked you in secret channels :P [16:19] that change could only affect camera/video in theory [16:19] yeah [16:19] and I tested that before uploading it, but let me check [16:20] but by the looks of it people dont have the container running [16:20] http://people.canonical.com/~ogra/touch-image-stats/150.changes ... [16:20] downloading it now [16:20] lxc only was a no-change rebuild [16:20] new packages init? [16:20] and i dont see what else could have broken it [16:20] yeah, thats a meta package, wont do anything [16:20] alright [16:21] well, let me flash and see :-) [16:21] right, running an OTA now too [16:21] fginther, care to give your mp to land reminders a whirl again? https://code.launchpad.net/~fginther/ubuntu-test-cases/add-reminders/+merge/226281 [16:22] * sil2100 will wait with sending out the e-mail once we know more [16:23] Curse you jenkiiins! [16:23] balloons, sure [16:24] kenvandine: heya..so, whats your expected order of landings for ubuntu-system-settings ? [16:24] http://people.canonical.com/~platform/citrain_dashboard/#?q=ubuntu-system-settings [16:24] hmm, my OTA try doesnt move its progressbar at all [16:24] fginther, note, there's a branch you should use that is pending to land in trunk [16:25] if it works, I'll land it and we'll go ;-) [16:28] sil2100, bootstrapped is fine [16:30] ogra_: flashed 150 on flo, with bootstrap, working fine [16:30] let me flash 149 and try ota [16:30] rsalveti, yep - same here [16:30] rsalveti, doesnt start for me [16:30] rsalveti, but worrying that update doesn't work [16:30] so it is ota that is killing it then [16:30] right, might be an issue with the update [16:30] ogra_, from bootstrap? [16:30] and i cant get in via pablet shell anymore, i guess mterrys dropping of the password broke that [16:31] so does that mean that the url dispatcher bug my be relevant? [16:31] brendand, nope, OTA from a 149 bootstrap [16:31] s/bug/crash [16:31] ogra_: adb shell worked for me [16:31] kgunn, once the current landing is merged i'll let you guys have a slot to land, ok? [16:31] clean 150 [16:31] hmmm [16:31] sil2100, i guess its slangasek/mmterry changing the password handling that breaks [16:31] kgunn, i assume you have one ready, just need to merge right? [16:31] rsalveti, phablet-shell doesnt anymore [16:32] sil2100, thats not upgrade safe [16:32] ogra_, i got in by phablet-shell, from bootstrap [16:32] right [16:32] that also explains why fresh flashing works [16:32] Where did that change? Didn't it change earlier, in 149? [16:32] brendand, right [16:32] kenvandine: oh, no prob...was wondering if there was a preferred order to things [16:32] sil2100, livecd-rootfs [16:32] * sil2100 remembers hearing about it in the morning image [16:33] kgunn, not really... just we had that one all ready to land [16:33] ogra_: I adb shelled in on 150 with ota [16:33] ogra_: but why would that break ota? [16:33] * ogra_ hugs kenvandine [16:33] Ok [16:33] * kenvandine hugs ogra_ [16:33] kenvandine, thanks ! lots of beer for you next time we meet :) [16:33] davmor2, adb works fine, as i said [16:33] :-D [16:33] davmor2, phablet-shell doesnt [16:33] :) [16:33] d'oh misread sorry [16:34] rsalveti, because passwd expects a user in /var/lib/extrausers/passwd [16:34] * sil2100 checks the livecd-rootfs change [16:34] rsalveti, whgich is created *only* by livecd-rootfs [16:34] right, but we're also updating the rootfs [16:34] rsalveti, but not when upgrading [16:34] so whatever gets done by livecd-rootfs, gets also applied on upgrades [16:34] why not? [16:34] rsalveti, its a writabel file ;) [16:34] oh [16:34] bad bad [16:35] yeah [16:35] Ouch [16:35] nobody thought about that i guess [16:35] and neither slangasek nor mterry seem to be around [16:35] * sil2100 pokes slangasek [16:35] ogra_: bet they are thinking about it now though [16:35] erm, I'm certainly around [16:35] ah, hey :) [16:36] slangasek, upgrades broke ... we dont migrate the phablet user to libnss-extrausers on upgrades [16:36] which leaves us completely without any user after upgrading [16:36] how can that be? Shouldn't that be automatically handled by system-image? [16:36] (freshly flashing works fine) [16:36] does the delta on the channel not show the /var/lib/ files in the tarball? [16:37] slangasek, i guess the prob is that the extrausers password file is writable ... [16:37] we dotn really have a way to handle upgrades of writable files [16:37] maybe needs a hack in lxc-android-config or something [16:37] root@ubuntu-phablet:~# ls /var/lib/extrausers/ [16:37] root@ubuntu-phablet:~# [16:37] hum [16:37] cjwatson, yeah, something like that [16:37] then one wonders what good it's supposed to do to add adduser support at all [16:38] these are unrealted [16:38] we simply need some lxc-android-config magic to move existing users over [16:38] no, they're not [16:38] since upgrades != fresh installs [16:39] or is 'adduser' expected to be run only for local accounts, post-install? [16:39] if that's the case, then ok, not a problem [16:40] adduser is used to a) create during build ... (thats the bit the "hack" works around), but also to add/remove the pahblet user from groups etc [16:40] i think nobody took into account that it will break during upgrades so thats a new scenario ... [16:41] and i dont really get why i have no passwd file *at all* under /var/lib/extrausers [16:41] do we make the whole dir writable ? [16:41] instead of the three files [16:42] root@ubuntu-phablet:~# grep extrausers /etc/system-image/writable-paths [16:42] /var/lib/extrausers auto persistent transition none [16:42] aha ! [16:42] hmm, shouldnt transition use the existing files in there ? [16:42] ok, I'm going to assign bfiller's silo out of sequence because (a) selfish and (b) it should be pretty fast to land [16:42] root@ubuntu-phablet:~# umount /var/lib/extrausers [16:42] we should get silo 4 freeing up soonish [16:42] root@ubuntu-phablet:~# ls /var/lib/extrausers [16:42] group passwd shadow [16:42] aha [16:44] * ogra_ tries to list all the files separately [16:44] yup [16:44] fixes it [16:45] let me upload a fix ... [16:45] dch warning: Recognised distributions are: [16:45] {hardy,lucid,maverick,natty,oneiric,precise,quantal}{,-updates,-security,-proposed,-backports} and UNRELEASED. [16:45] is citrain jenkins really using quantal? [16:45] it kind of ought to stop that [16:47] kenvandine: hey, feel free to go nuts and land away....i think silo6 we're gonna rebuild sunday night u.s. and land euro morning...and silo 12 is mterry's [16:47] he may or may not land today...most likely monday [16:49] slangasek, http://paste.ubuntu.com/7857314/ === gatox_lunch is now known as gatox [16:49] ogra_, which landing broke it then? [16:49] brendand, livecd-rootfs [16:50] ogra_: typo, "passw" [16:50] well, actually lxc-android-config weeks ago := [16:50] cjwatson, oops [16:50] +/var/lib/extrausers/passw auto persistent transition none [16:50] ogra_, could we have caught it without actually doing an upgrade? [16:50] brendand, well, we only started making use of the feature with 150 [16:50] cjwatson, right, fixed locally, thanks [16:50] (also typo "pahblet", but that's only in the changelog ...) [16:51] yw [16:51] http://paste.ubuntu.com/7857334/ [16:51] yeah, fixing the changelog too [16:51] * cjwatson prods silo 14, which apparently I missed earlier [16:52] sil2100, i'll do a dircet upload of lxc-android-config for this now ... 151 should be fine again then [16:52] *direct [16:52] * ogra_ uploads [16:53] ogra_: ok, thanks! [16:53] ogra_: seems like a sensible fix [16:53] nah, the bug is that "transition" doesnt do what it should [16:53] but its late on a friday ... nothing to inspect now :) [16:53] ;) [16:54] tehoretically "transition" should have copied the directory content to the bind mounted tmpfs [16:54] ogra_: could you make sure an image is kicked of as soon as this lands in the archive? [16:54] As this is serious and we cannot allow waiting [16:55] sil2100, yes, though i have to go out for a bit, might be slightly delayed due to that [16:55] Ok, I'm also AFKing for some hours but then will be back [16:55] I should be around to kick it in [16:55] robru: could you also look out for this ^ ? [16:55] slangasek: thanks :) [16:56] sil2100: robru: hi, could you guys please reconfigure silo 001? there are some new components in there [16:56] ok [16:59] * sil2100 out [16:59] robru: thanks [17:00] boiko, you're welcome [17:01] ogra_: man next you'll be saying you have a life and you go out on Fridays with "Real" people ;) [17:01] people ? [17:01] * ogra_ looks up on wikipedia [17:01] hahaha [17:02] tea time === alan_g is now known as alan_g|EOW [17:55] sil2100, bah, looks like gallery-app regressed [17:55] * ogra_ suspects TRAINCON-0 gets more likely every minute [17:56] 3 errors in smoke testing [17:57] ugh ... shorts too [18:00] sigh ... and dialer-app [18:03] ogra_, on 150? [18:03] pmcgowan, yes http://ci.ubuntu.com/smokeng/utopic/touch/mako/150:20140725.1:20140725.1/9255/ [18:04] still running (all the 0% ones) [18:04] why would those tests start failing, I guess thats always the question [18:06] right [18:06] there is no apparentl reason on http://people.canonical.com/~ogra/touch-image-stats/150.changes [18:21] slangasek, lxc-android-config migrated ... want to kick the image ? [18:22] I'll do it now [18:22] * ogra_ is afk now [18:29] ogra_: minute, milisecond ;) [18:29] ogra_: hows the fix doing [18:29] === trainguard: IMAGE 151 building (started: 20140725 18:30) === [18:29] \o/ [18:30] cjwatson: thanks [18:40] slangasek: here is error from silo https://ci-train.ubuntu.com/job/landing-001-1-build/133/console [18:40] slangasek: maybe we just need to force a rebuild, wasn't sure [18:40] bfiller: "You need to ensure that your version contains the fix in the destination or you can force rebuild to bypass the check" - well, we want to force and bypass here :) [18:41] slangasek: ack, just wanted to make sure that was the right thing to do [18:41] yep, thanks for checking [18:52] biab need to collect wifey === cprov changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): cprov | CI Train support: cjwatson | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [19:19] fginther: are silos buildable now? [19:21] ogra_: how can you be afk and type at the same time? dictaphone? [19:24] sergiusens, should be, the chroot problem from a few hours ago has been resolved and a number of builds have succeeded [19:35] Oh come on 151 [19:35] ogra_: the gallery-app smoketest regressions all relate to writing pictures to /home/phablet/Pictures. Any possibility that this is related to the nss changes, you think? [19:37] slangasek: could be 151 should be able to confirm it though [19:43] we're in the process of landing a new release on ci.ubuntu.com and have disabled updates while this takes place. ETA 1 hour. [19:47] 151 - our hope ;) [19:54] sil2100: yeah but it isn't built yet and it's way past my EOW ;) [19:54] davmor2: WHAT?! [19:54] davmor2: how DARE you have a weekend ;) [19:55] davmor2: how DARE you have a life! [19:55] ;p [19:55] sil2100: It's just 2 days to dogfood rather than a couple of hour but don't let on to the wife she thinks I'm just playing :D [19:57] hah ;) [19:57] Well, anyway, we'll promote something on Monday [19:57] So no need to flex ourselves [19:57] sil2100, yeah, promote no matter how broken it is ;-) [19:57] sil2100: no I just need a bloody phone for the weekend and want to test that the fix fixes [19:59] === trainguard: IMAGE 151 DONE (finished: 20140725 20:00) === [19:59] === changelog: http://people.canonical.com/~ogra/touch-image-stats/151.changes === [19:59] \o/ [19:59] \p/ [20:00] flashing... now! [20:00] Now tell us if it works! [20:01] * sil2100 writes the e-mail in the background [20:07] flashing............ [20:08] waiting for the spinny logo with the line under it to go away [20:11] sil2100: I haz a phone [20:11] now to quickly see if it works [20:12] seems to [20:12] \o/ [20:13] sil2100: I use system-image-cli -b 0 to flash and it takes a while [20:13] via adb shell [20:15] Ok, I'll go take a shower and then finish up the e-mail [20:16] robru: need another reconfigure on silo 1 please [20:17] bfiller, done: https://ci-train.ubuntu.com/job/prepare-silo/1167/console [20:17] robru: cheers [20:17] bfiller, you're welcome [20:18] sil2100, davmor2: well 151 seems to boot properly at least [20:19] robru: yeah I've added account sent an sms and browsed the web so it seems to work [20:19] great [20:19] PROMOTE IT! ;-) [20:19] haha [20:19] great, working image again :-) [20:29] Right night all catch you on Monday Traincon0 here we come [20:32] fginther, hey, so I just merged & deployed your chroot update script, all looks good there. I'd like to run it just to see for sure it's working, will that disrupt existing builds? or can the system handle chroots being updated out from underneath them? [20:33] robru, it's safe to run the upgrade at any time [20:33] fginther, sweet, thanks [20:35] fginther, score, you rock [20:36] robru, glad it's working [20:36] fginther, yeah me too ;-) [20:36] robru, this should resolve the issue that slangasek brought up earlier this week. [20:37] fginther, yeah, for sure. I had been meaning to look into this, but I'm really glad you beat me to it ;-) [20:37] huzzah [20:37] robru, leave it to a disaster to provide some motivation :-) [20:37] that should substantially cut down on the silo deploy time [20:37] fginther, oh yeah, what was the disaster? were they just so stale that builds were failing? [20:39] robru, it was the same broken upgrade path (the one that required the apt-get dist-upgrade / -f install dance) that was also causing the builds to break. [20:40] fginther, ah yeah, makes sense. [20:42] cjwatson: robru: seems credentials are expired? https://ci-train.ubuntu.com/job/landing-008-0-reconfigure/33/console [20:42] cjwatson: are you on train call all day? [20:42] sergiusens, oops, no, it's my turn [20:42] :-) [20:43] sergiusens, connection timed out looks transient, try it again? === robru changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): cprov | CI Train support: robru, trainguards | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [20:47] ci.ubuntu.com update is now complete [20:55] robru: I'm getting same timeout https://ci-train.ubuntu.com/job/landing-009-0-reconfigure/31/console [20:55] tried a few times [20:56] robru: ack [20:57] ugh same problem with kenvandine 's publish job [20:57] yup [20:57] :/ [20:57] it must be a friday afternoon! [21:00] must be! [21:00] robru, should i try again? [21:00] oh, it looks like when I deployed fginther's chroot fix, there was some garbage in trunk that also got deployed. gimme a minute to revert it [21:00] robru, ok, thx === cprov changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard (general help): cihelp | CI Train support: robru, trainguards | CI Train Status: #133 promoted | Known issues: Both queuebot and http://bit.ly/1mDv1FS know your silo status before the spreadsheet does. [21:07] robru: got the error again [21:08] robru: got the error again [21:08] sergiusens, yep, working on it [21:08] robru: got the error again [21:08] sorry if multiple pings reach you [21:08] network issues [21:08] not being a PITA :-) [21:08] sergiusens, no worries [21:12] sergiusens, bfiller kenvandine : ok everybody, try it all again now [21:15] this poped into my mind https://www.youtube.com/watch?v=LtjhW-ERoak [21:15] except the i love yout part :-P [21:24] kenvandine, published for you, looking good now [21:45] barry, still around? [21:45] robru: yep [21:45] barry, I'm working through your test plan. seems ok so far. i'm hoping to complete the plan within the hour [21:46] robru: awesome. i wonder if i should violate warsaw's 2nd law and publish it today or wait until after traincon 0 on monday [21:47] barry, i'll let you know when I'm done testing ;-) [21:47] robru: fair enough :) [21:50] barry, so in test B I'm finding that it claims to be installing, but the progress meter never moves beyond 0%. is that normal? [21:51] barry, like, I just hit pause and then it jumped to 97%. [21:52] robru: it should act pretty much the same as test plan B. pmcgowen over in #u-t had other problems with B [21:52] of course it wfm [21:52] robru: now that 151 is out, i will try test B again here [21:53] barry, yeah, it seems to do the install & reboot just fine, just that the download does not actually report any progress while it's happenin [21:54] robru: weird [21:54] barry, yep [21:55] it's a rather odd test, but i'm not sure what you describe would be enough to hold up a release. let me see what happens for me [21:55] barry, I accidentally did test plan b once without actually installing 2.3.1 and got the same behavior, so it's not a new regression (at least it's not a regression ins ystem-image, might be a regression in 151) [21:56] robru: is there a new u-d-m in 151? [21:57] barry, no, but there is system-settings http://people.canonical.com/~ogra/touch-image-stats/151.changes [21:57] that's interesting [21:57] that could certainly have an effect [21:58] robru, thx! [21:59] kenvandine, you're welcome! [22:00] robru, can you merge and clean silo 2 when it's ready? [22:09] robru: progress worked fine for me. one thing to check is to make sure system-image-dbus isn't already running before you open the systems settings panel. apparently s-i-dbus now gets started at boot time (dunno why) [22:10] barry, that seems like a bug though, right? I mean you can't expect users to know to kill that before checking for updates. [22:11] robru: test b is somewhat contrived though. users aren't going to touch their channel.ini file [22:15] true [22:16] robru: yeah, it works okay for me. publish or wait? [22:16] barry, sorry I got distracted by some stuff. just about to do test c [22:16] barry, unless you think my approval of test a is good enough to publish on? [22:16] because test a was fine [22:17] robru: i can wait for you to do test c - that's a manual download so would be useful to verify [22:17] barry, ok, one sec [22:23] barry, so far so good, just waiting for the update to finish installing [22:25] barry, yep, test c a success, let's publish ;-) [22:26] barry, just need you to approve the MP [22:34] robru: \o/ [22:35] robru: hmm. status only allows me to set it to wip or merged [22:36] ah commit message [22:37] robru: hmm, no that's not it [22:37] what do you see under status: https://code.launchpad.net/~barry/ubuntu-system-image/systemimage2.3.1/+merge/228010 [22:39] "lp:~ubuntu-managed-branches/"? [22:39] slangasek: that's supposed to be the branch to merge into [22:40] well, but what is that branch? [22:40] it's an unfamiliar one to me [22:40] slangasek: it's the citrain branch that stgraber set up [22:40] ah, ok [22:41] and i am not a member of that team. only stgraber and the ps jenkins bot is [22:41] :/ [22:42] i bet stgraber is eow. in which case, well i guess it'll have to wait until monday! [22:42] barry, why is the mp pointint at ubuntu-managed-branches team? [22:42] barry, like, why isn't there a real team for this? [22:43] stgraber was eow on Monday [22:43] but will be back next week [22:43] robru: we develop system-image like a good upstream should :) [22:43] slangasek: ok. that's probably a sign not to violate warsaw's 2nd law anyway [22:43] barry: I'm not sure that follows. the other upstreams have a trunk branch that people can commit to. [22:43] barry, oh, well ps-jenkins is in the team, so I can publish if you want. [22:44] just have to override the merge approval thing [22:44] robru: if you can, go for it [22:44] barry, no, this is weird. you can't even approve your own merges. [22:44] barry, i mean yeah, I'm gonna publish, but I wouldn't describe this scenario as what "good upstreams should do" [22:44] robru: because i'm not in the team that owns the branch my branch is merging into [22:45] meaning: s-i is an 'upstream project' in the sense that it's not a native package [22:45] barry, but citrain defaults to not publishing unapproved merges. you've put yourself in a situation where stgraber is literally the only person who can approve your landings. [22:45] robru: i think that's a historical mistake [22:46] not sure why we didn't notice this before, but things were different back in march [22:47] barry, yeah, the thing that disallows publishing unapproved merges is a bit newer. [22:48] bfiller: thanks for tracking down my gallery-app bug [22:48] barry, it's ok, I overrode it, but it's just like, now any time we want to do s-i landings, we have to just know to override that. it's breaking the default assumption of citrain that upstreams can approve their own merges. [22:49] robru: yep. i just requested from stgraber to add me to that team. so it should all be good when he gets back on monday [22:49] barry, I don't understand the virtue of not wanting to be able to approve your own merges. that doesn't sound like upstream at all, that sounds like you letting the distro be the gatekeeper of your work. in citrain we consider "upstreams" as independent teams who control their own trunks [22:49] barry, ok no worries I guess. [22:50] robru: right. i control my own trunk, which is lp:ubuntu-system-image. no worries about pushing, mp'ing, approving to that. the ~u-m-b branch is the "debianized" packaging branch only. no upstream work happens there, just landings [22:50] robru: so you're probably right that we didn't hit this before because the gate feature is newer than our last landing [22:52] we worked all this out at the london sprint, but weren't caught by this new feature until now [22:52] I ditched the ubuntu-managed-branches thing for click a while back after deciding it made no sense [22:52] cjwatson: np [22:52] robru: in any case, thanks for publishing it [22:52] I just use lp:click and lp:click/devel now [22:53] cjwatson, yeah, heh, I didn't even know there was an ubuntu-managed-branches thing until just now [22:53] barry, you're welcome! [22:53] cjwatson: at the time, i really didn't want to go through the train for every commit, just for new version landings, so this seemed to make the most sense. maybe now that the train is more streamlined, it doesn't any more [22:54] and for some reason, i'm still resistant to making lp:system-image a native package :/ [22:55] barry, doesn't need to be native... you mean inlined? [22:55] barry: I don't go through the train for every commit [22:55] robru: not sure if this is a bug, but lately in citrain device-upgrade it does the install but then I have to run apt-get -f install on the device as it reports the packages are not fully installed [22:55] robru: i don't want debian/ i my trunk [22:55] robru: I may have an old version too.. [22:55] *in my trunk [22:56] barry, right, that's inlining, not going native. the default thing in citrain is that debian/ is in your trunk, but it's a split package, not native. [22:56] cjwatson: hmm. yeah, maybe i'm just being dumb [22:56] bfiller, yeah, something or other broke in launchpad and now my citrain script can't find the packages from the ppas anymore. cjwatson do you know if that's ever going to get fixed in launchpad? [22:57] robru: what? [22:57] I might need slightly more details there :) [22:57] cjwatson, i asked you about this a while ago, the 'devel' symlink is gone from ppas [22:57] uh [22:57] oh that's the htaccess thing [22:58] cjwatson, that'd be super if that could get fixed ;-) [22:58] yeah I think that was going to be fixed next week, it just needs to be coordinated with a code rollout [22:58] William was sorting that out [22:58] saw ops chatter earlier today about it [22:59] cjwatson, oh yeah? good chatter or bad? [22:59] good chatter [22:59] ok cool [22:59] it was all part of fixing PPAs to be able to handle non-Ubuntu [22:59] bfiller, ^ should get fixed next week apparently [22:59] which is critical path for RTM [22:59] robru: nice [23:00] it's obviously trivial to work around temporarily in your script though [23:00] I mean even just locally if you don't want to have to work out when you need trusty and when you need utopic [23:01] cp -a /usr/bin/citrain ~/ && sed -i s,/devel/,/utopic/, ~/citrain [23:19] bfiller, ^^ you can try that ;-) [23:19] cjwatson, the thing is I don't want to do a whole release for that and then revert that with another release if devel is just coming back anyway [23:22] robru: sure, not suggesting you do it, suggesting that it's easy for anyone affected by it to temporarily do it locally [23:22] hence the form of the example I gave [23:23] cjwatson, yeah, thanks === Ursinha_ is now known as Ursinha === cjohnston_ is now known as cjohnston [23:54] alright, I'm out. gotta be up early for guadec tomorrow! [23:56] xnox: https://wiki.ubuntu.com/Chipaca/PPU [23:56] Chipaca: tah.