[00:40] ev, not sure of the actual load these days, but the calxeda's that came in sure made it a lot less dramatic for the ex-PS stuff; I do want to say that we use to have a dput per merge into PPAs (binary version of trunk) and were told to stop doing that because we were overloading the system; again this is prior to calxeda [00:42] if resource wise is ok, all that would need to be devised is a way to pull out the test results; not that it would be difficult [01:03] infinity: rmadison reports new mir is in, didrocks suggested to ask you to kick off a new image build? [01:04] cyphermox: FYI ^ [01:06] ogra_: we can kick an new image (in case infinity can't get it before you are online). The Mir fix is in [01:06] ogra_: infinity: is there any more to that than running the line from crontab? [01:06] (I could do it) [01:14] Laney: Should be just that. Sorry, was in transit between hotels. [01:14] mmkay [01:14] Laney: You already doing it, or should I? [01:15] infinity: ok, Laney volonteered :) [01:19] voluntold [01:21] ev: We could look at the images on system-image and cdimage for better numbers, but it's generall 1-2 builds/day. We didn't have one today though [03:18] in case anyone was waiting around for it, image showed up and tests have started on it - the image ci shouldn't truncate .crash files anymore too [03:39] asac, ogra_, rsalveti: gallery passes again \o/ [03:58] plars: :D [08:09] plars, yay [08:25] cyphermox, your latest upload of autopilot 1.3 breaks system upgrade if python-autopilot-strace is installed bug 1246620 [08:25] bug 1246620 in ust (Ubuntu) "upgrade failed: trying to overwrite '/usr/lib/x86_64-linux-gnu/liblttng-ust-tracepoint.so.0.0.0', which is also in package liblttng-ust0:amd64 2.1.1-6" [High,Triaged] https://launchpad.net/bugs/1246620 [08:30] jibel: got the same issue yesterday when upgrading my desktop [08:30] cyphermox: ^ [08:43] fginther: hello, poke me back when you're around :) === psivaa changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: psivaa | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [10:07] psivaa, ogra_: did you see that ubuntu-ui-toolkit tests are all failing now? [10:08] plars: uitoolkit passed in image 9 iirc [10:08] no, i noticed that maguro had not run gallery when i looked last [10:09] (but thats a while ago) [10:09] i re-ran gallery and it passed [10:09] great [10:09] psivaa, ogra_: *sigh* sorry, nevermind me... I've been up since 3am de-flooding my house [10:09] * psivaa checks the time in texas [10:10] psivaa: it's 5am now [10:10] plars, urgh ! [10:10] psivaa: I saw you restarted unity8 on mako, thanks - it looks promising from what I saw on maguro [10:11] plars: yea, just reran a few jobs. the passrates are actually good with 9 [10:11] would it be a good idea not to accept a new image if it doesn't pass UITK (trunk) autopilot tests? [10:11] psivaa: I saw, awesome [10:11] ogra_: psivaa reran that one already on maguro [10:11] ogra_: it was good [10:11] and with "accept a new image" I mean make it the new default image for CI [10:11] plars: btw good luck with de-flooding and having a bit of sleep :) [10:12] ok, I'm going to go at least lay down for a bit, I think sleep is out of the question at this point [10:14] oh man [10:14] good luck [10:14] psivaa, do we know why unit8 fails on mako ? smells like not unlocked or some such [10:15] ogra_: not really, we have a better passrate on maguro. unfortunately i dont have a mako to locally test [10:16] looking through the logs to see anything obvious [10:20] ogra_: we have unity8 crash there [10:21] we do on maguro as well though [10:21] two even [10:29] yea, lauch_unity is failing, tried looking at the network camera and can't see any movements. that may be pointing that the screen is locked [10:31] yup [11:16] ogra_: no new build yet? [11:17] hmm ? we have #9 [11:17] was built at 3am UTC [11:17] (and still didnt pass the unity8 test on mako) [11:26] untiy8 on mako was tried again after re-flashing, still dint pass. [11:32] :( [11:40] oh, sorry, i thought #9 was the old broken one, not the new broken one ㋛ [11:41] popey: No 8 was the old broken one :) [11:41] popey: at least apps stay open long enough to do things now :) [11:50] right, 9 looks pretty good ... except that unity8 thing :/ [11:58] ogra_: would you recommend upgrading my test phone to #9 or is it unusable? [11:58] popey, looks fine on maguro is all i can say [11:59] i wont upgrade my mako now that it replaces my android phone :) === cjohnston changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: cjohnston | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [12:01] * popey updates [12:01] ogra_: next time you use update manager and see new apps, before you install them can you please see if you can reproduce bug 1245890 [12:01] bug 1245890 in Unity 8 "Pinning to launcher followed by upgrade, breaks pinned app link" [Undecided,New] https://launchpad.net/bugs/1245890 [12:02] popey, cant confirm, this works for me [12:02] * ogra_ usually adds his new apps in the stroe to the launcher ... they all still work fine [12:19] hm [12:20] popey, that's a dup [12:22] * ogra_ shakes his fist at google === alan_g is now known as alan_g|lunch [13:17] fginther: hello! Re-ping ;) [13:17] fginther: did you see my annoying PM? [13:19] sil2100, sorry, just saw it [13:19] sil2100, one moment [13:20] Thanks! === alan_g|lunch is now known as alan_g [13:39] hi guys, I need a new release of lp:history-service, how should I proceed? [13:41] boiko: http://paste.ubuntu.com/6292280/ [13:42] boiko: hi! [13:44] cjohnston: hi, so, I have seen those instructions, but there is no link to the Landing Asks sheet [13:44] sil2100: hi! [13:44] boiko: as it's history-service we're talking about, I would proceed by adding it to the Landing Asks [13:44] boiko: let me give it to you, but you should actually ask your manager to do that [13:44] yeah, unlikely you can edit it [13:44] boiko: https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0Au6idq7TkpUUdGNWb0tTVmJLVzFZd0doV3dVOGpWemc#gid=1 [13:45] what sil2100 said [13:45] after all it is still the same process we use since twoo months though [13:46] boiko: we can then follow up and get it released in the nearest time - since didrocks is on a sprint I guess it might take till next week to get it released? Depends if the guys in Oakland pick it up or not [13:46] sil2100: oh right, thanks, I am just asking directly cause everybody is at the sprint [13:46] popey, did you try #9 on mako yet ? [13:47] boiko: I would try releasing it myself, but I would first need management permission, and they're at the sprint as well... [13:47] * ogra_ wonders if unity8 is fine in real life ... compared to the failing tests [13:47] boiko: what issues does it fix? [13:47] ogra_: I can upgrade my device if needed [13:48] sil2100, why do you need management approval ? if the AP tests pass and you cant see a regression in a manual smoke test it shouldnt be blocked on this [13:48] sil2100: well, the one bit I need released is the pkgconfig support, but there has been other changes there as well (things that were pending from the last cycle) [13:48] ogra_: I didn't know if that's ok right now - but if you say that, then I guess! [13:49] well, we shouldnt block on them not being here ... as long as the landing appears safe [13:49] * sergiusens thinks that when managers go on sprint they should get some peer+1 assigned [13:49] (test twice if you feel thats safer :) ) [13:49] ;) [13:49] In my case even testing thrice won't help [13:49] :D [13:50] Anyway, thanks, then I'll look into it and try getting it released [13:50] well, if it doesnt pass, indeed, keep it til next week [13:50] but things that pass safely shouldnt be blocked by missing managers [13:53] sil2100: I can ask my manager if that makes you more comfortable, that's fine, the landing is not urgent [14:06] ogra_: yeah, it's been fine here [14:06] thought so [14:28] plars: excellent news on no longer truncating crash files [14:28] on the frequency of builds, I was referring more to how often we're running dpkg-buildpackage within the CI infrastructure === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [15:56] jibel: well, it's not a breakage in autopilot. === alan_g is now known as alan_g|tea [15:59] jibel: that said, it's on my list, I'll fix it shortly [15:59] sil2100: ^ [15:59] cyphermox: thanks! === doanac changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: doanac | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [16:04] doanac, hey ... [16:04] ogra_: what's up? [16:04] doanac, it looks like we have issues with unlocking the screen on mako with the unity8 tests ... [16:05] psivaa ran them multiple times on image #9 ... he checked with the webcam and apparently the screen doesnt show anything ... popey tested #9 manually and unity8 behaves fine === alan_g|tea is now known as alan_g [16:06] ogra_: i don't think we run unlock_screen for unity8. we just kill the shell before running. [16:06] checking now [16:06] hmm [16:06] doanac: the issue is only on mako, we are though not 100% but fairly well in maguro [16:07] well, there seems to not be anything unusual in the logs apparently [16:07] doanac: lauch_unity is failing [16:08] let me test locally and see what happens [16:14] sil2100: are the cu2d problems because of autopilot or what's all the red about? [16:15] sil2100: there seem to have been also multiple manual uploads I guess (yellow) [16:15] ogra_: i'm running the test at home and its failing the same way [16:16] wow [16:16] i wonder what makes mako special then [16:17] Mirv: autopilot was reddish today as well, so I would say it's fifty-fifty - I was mostly dealing with other things today so I didn't administer cu2d in detail [16:18] sil2100: it's not autopilot tests, the check job looks problematic in general somehow http://10.97.0.1:8080/job/autopilot-trusty-daily_release/281/label=autopilot-nvidia/console [16:18] lxc problem? [16:19] Mirv: I was using the autopilot jobs directly today, without cu2d, and just saw that there was a loong stroke of red jobs, but didn't have a moment to look into that [16:20] Let me also look at that now [16:20] Mirv: ah, this one - this one I saw happening once, but then next time the machine ran fine [16:20] sil2100: it seems it's nvidia specific [16:20] Yes, let me find my job, one moment [16:21] then under qa there is also a missing package when looking at the intel side [16:21] Aaah [16:21] Mirv: ok, my mistake, it was happening all the time on nvidia, so it seems to be b0rken [16:22] sil2100: can you look at that still, you have much more skillz in the lxc area? :) [16:22] I'm fixing at least qa package list now [16:22] Ok, let me take a look then [16:24] my main problem is that if it complains about a missing container, I wouldn't know immediately where I should see the containers etc. [16:26] hmm, I might see one of the problems [16:27] The permissions to the containers directory got changed, not sure if jenkins user can access them this way - so it cannot find them [16:27] Let me try fixing that [16:28] fginther: the permissions thing ^^^ - weren't you seeing something similar the other day? [16:29] Mirv: let's run some test now maybe? [16:29] cyphermox: synced your autopilot manual upload to the branch to make the prepare job work again [16:30] sil2100: aha.. ok, I'll try [16:32] sil2100: QA running [16:32] sil2100: actually there's another check ongoing, nvidia seems to be running [16:32] let's wait [16:34] Mirv: keeping my fingers crossed! [16:34] * sil2100 wonders if that's the issue [16:34] We'll see soon enough [16:34] thanks [16:36] sil2100, Mirv this is a change in lxc 1.0.0~alpha2-0ubuntu5, you can revert that change by chmod'ing /var/lib/lxc to make it readable by any user. [16:37] sil2100, Mirv next upgrade won't touch it [16:37] jibel: so I looked at NM again, hard, and I can't figure anythign wrong with dnsmasq... would it be possible to have the autopkgtest run again so I could see what goes on? [16:38] jibel: did that [16:38] cyphermox, sure, but it ran again this morning and I had to kill it again. Didn't pitti found the reason? [16:38] jibel: that's what I'm trying to test now :) [16:39] jibel: cool, thans [16:39] jibel: oh, you did [16:39] But good, then I diagnosed it correctly, thanks [16:39] sil2100, at least that what is written in the changelog :) [16:39] jibel: alright then [16:40] jibel: I haven't heard back from piiti about it.. [16:40] jibel: was the environment build from a server image or from minimal? [16:41] doanac, where? Yes we saw an issue with the workspace dir getting set to 600 causing jenkins to complain of mkdir failures [17:01] cyphermox, it's based on a server cloud-image http://developer.ubuntu.com/packaging/html/auto-pkg-test.html#executing-the-test is you want to test locally [17:02] weee [17:02] awesome, that's what I expected. [17:02] I'll definitely give it a good look, I think I know what the issue is [17:04] Mirv: did it help? [17:04] I need to disconnect now, but I'll be back in like 3-4 hours [17:04] Mirv: drop me an e-mail if anything ;) [17:05] sil2100: yes it helped [17:05] at least initially it looks so [17:05] \o/ === Ursinha is now known as Ursinha-afk [17:10] ok soon many of the stacks should be greener/yellower, as I'm rerunning check jobs for a bunch of them [17:10] fginther: I've seen some Mir jobs today with "E: Unable to locate package liblttng-ust0" e.g. https://jenkins.qa.ubuntu.com/job/mir-android-trusty-i386-build/73/console - can you help? [17:10] build failures in some too, though [17:13] Mirv: https://code.launchpad.net/~mathieu-tl/indicator-datetime/13.10.0+13.10.20131023.2-0ubuntu1/+merge/193460 [17:16] +1 === Ursinha-afk is now known as Ursinha [17:37] jibel: what hoops do I need to go through if I want to delete that offending test? [17:37] it's just too broken, not bringing much usefulness [17:37] (it's running fine on my machine with the auto testing scripts) [17:41] alan_g: mir should've been rebuilt against liblttng-ust2 recently, I thought [17:42] alan_g: looks that way in trusty anyway. maybe transient? [17:44] cjwatson: what was the issue? [17:52] cyphermox: 17:10 fginther: I've seen some Mir jobs today with "E: Unable to locate package liblttng-ust0" e.g. https://jenkins.qa.ubuntu.com/job/mir-android-trusty-i386-build/73/console - can you help? [17:52] cjwatson, thanks for the reping [17:52] cjwatson: Hmm... I guess the https://jenkins.qa.ubuntu.com/job/mir-android-trusty-i386-build/73/console runs the tools/setup-partial-armhf-chroot.sh script? [17:53] ah, yeah [17:54] and that script needs to pull liblttng-ust2? [17:55] that should just be happening automatically, python-autopilot-trace depends on it [17:55] alan_g: That's beyond my knowledge, but I would've hoped it'd be following mir's dependencies rather than hardcoding a library name [17:55] cyphermox: Downloading mir dependency: liblttng-ust0 [17:55] oh, mir [17:55] I think it's probably worth retrying to confirm it's gone [17:55] Which it should be ... [17:56] cjwatson: that's an embarrassing script, but no-one has spent the time to fix it properly [17:59] ah [17:59] yeah, that's hard-coded there === plars changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: plars | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [18:00] hardcoded> yikes! [18:03] :q [18:03] alan_g, cjohnston, cyphermox, so problem solved? setup-partial-armhf-chroot.sh needs to be updated? [18:03] yeah. updated with rm if possible ;) [18:03] fginther: otherwise I'd like to fix it to get the right stuff from debian/control [18:04] cyphermox, I'll leave that to the mir team [18:04] fginther: yeah - I'm at EOD. I've passed the problem to kdub ;) [18:05] I'll propose a merge to repair this... I don't know what they use the script for really, but if it can't be dropped then.. === alan_g is now known as alan_g|EOD === Ursinha_ is now known as Ursinha === rsalveti_ is now known as rsalveti [19:00] fginther, could yo help me with this MR: https://code.launchpad.net/~om26er/mediaplayer-app/imrove_autopilot_tests/+merge/184584 [19:01] renato_: you should probably try the vanguard first [19:05] renato_: it looks like there is a test failure [19:08] renato_, the armhf node died during the build too [19:08] fginther, cjohnston I have approved again [19:09] fginther, renato_: https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-maguro/2813/console looks like unlock failed also [19:10] plars, the unlock worked on the second attempt [19:10] fginther: ah, ok I see [19:10] fginther: why is there no rebuild link on this one? do we just need to retry and post results manually if it passes? [19:11] plars: fginther: are you look at this one we'd need to launch a new image build? https://code.launchpad.net/~saviq/unity8/move-setenv/+merge/193426 (the e-mail thread "Unity 8 AP tests on #9") [19:11] plars, it was triggered by an approved MP, so to rebuild, you need to re-approve the mP [19:12] fginther: I have no rights to do that [19:12] fginther: do you? [19:12] renato_, there are a number of test failures, it may fail again [19:12] plars, yes, renato_ already re-approved [19:12] fginther, where did you see that [19:13] UNSTABLE: http://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/185 [19:13] renato_, something bad happened here too: FAILURE: http://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-maguro/2813/console [19:14] fginther, let me see if something change in the code since this mr was created because the first build works nice [19:14] renato_, the maguro failure looks like a possible crash [19:15] renato_, ahh, this was on build 8. that had some known issues [19:15] renato_, let me see if I can find the bug [19:16] Mirv: it seems to have restarted [19:16] Mirv: it's running now [19:18] renato_, https://bugs.launchpad.net/mir/+bug/1245958 [19:18] Ubuntu bug 1245958 in Mir "Apps crash with image 8" [Critical,Fix committed] [19:22] fginther, do you think this will work now? [19:22] bfiller, ^ ^ [19:23] renato_, the mako/maguro tests should be more stable [19:23] fginther: looks like the otto test failure is because correct codec not installed? https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/185/testReport/junit/mediaplayer_app.tests.test_player_with_video/TestPlayerWithVideo/test_show_controls_at_end_with_mouse_/ [19:25] bfiller, renato_, if that's the cause, then we should be able to resolve it by adding a dependency to the -autopilot package [19:25] rsalveti, told me to no do that [19:26] fginther, renato_ : I'll check with rsalveti - seems we need this, at least for otto [19:26] ack [19:27] plars: yeah, I just didn't do something so I'm wondering if you'd understand more of the problems it's seeing. [19:27] plars: I'm not holding my breath that it'd now succeed automatically [19:31] Mirv: it's not a failure I've personally seen before, did anything change? looks like it passed at one point in the mp [19:33] renato_: rsalveti said we shouldn't be testing mp4, and you should convert that to ogg or other free format [19:34] bfiller, ok I will try that [19:34] renato_: mp4 works on the device because it's using hardware decoding and doesn't need gstreamer-ugly, but on desktop that won't work without that pacakge. And you are right we don't want to install the -ugly [19:35] I am afraid that, the ogg does not work well on device (In the last time that I tried) [19:35] this was before the the new gst plugin [19:37] renato_: we might indeed have issues with a different format, but we need to use an open format [19:37] renato_: can you convert and post the video somewhere so we can test? [19:37] renato_: other we might indeed, temporarily, add the av/ffmpeg dependency for the test itself [19:38] *otherwise [19:43] rsalveti, give me a minute I will transcode the video [19:48] rsalveti, https://chinstrap.canonical.com/~renatofilho/videos/ [19:48] Mirv: well, it looks like it was mediumtests-trusty and qmluitests that failed last time, and they are passing now [19:48] could you try [19:48] Mirv: it's not 100% complete yet, but so far so good [20:00] Mirv: looks like it passed this time https://code.launchpad.net/~saviq/unity8/move-setenv/+merge/193426 [20:00] Saviq: ^ [20:09] plars: \o/ launching build [20:09] great [20:14] fginther, what is this error: https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-maguro/2877/console [20:33] renato_, looks like the device lost adb connection during the test. It's also at the same point in the test as https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-maguro/2813/console [20:33] renato_, is it possible the test itself is causing the device to reset? [20:34] fginther, I do not know, do you want me to try again: [20:36] I'll try another build of just the maguro job [20:54] * Mirv testing unity8 with the setenv fix [20:57] Mirv, plars, any reason why we dont just set that in the upstart job ? [20:58] ogra_: maybe a question for Saviq [20:59] yep, Saviq ^ [20:59] seems more appropriate for hacks ... also makes them more visible and saves you from changing the binary if you need to change or drop it [20:59] (not that this is overly important, just noticing it) [21:15] * Mirv published unity8 [21:15] ogra_: #10 after that, or do you know what's the plan? [21:30] ogra_, well, wherever they are - they shouldn't even be needed [21:46] unity8 now in release pocket === plars changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: - | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [22:11] Mirv, i'll trigger a build in the morning [22:11] * ogra_ is off for the night [23:14] Laney: infinity: can one of you kick a touch image build? [23:14] ok [23:14] there is an unity8 fix that we are interested in :) [23:14] thanks! [23:14] did the one I did yesterday work? [23:14] we should finally get a better pass rate [23:14] Laney: yeah, perfect! (but unity8 was crashing in the tests :p) [23:14] Laney: I won't blame you though ;) [23:14] ok, just wondering, you get no feedback [23:15] going [23:15] thanks! [23:21] how long does it usually take for the build? [23:24] popey: more than an hour [23:25] actually ~40 minutes for the ubuntu part [23:25] k [23:25] will have a play in the morning hten [23:25] then, I heard some 20 minutes for assembling the android one [23:25] popey: yep ;) [23:25] popey: still awake? getting late, isn't it? [23:26] yeah, watching politicians argue on telly before bed ☻ [23:28] some people count sheep