[01:15] sergiusens, sure [01:15] kgunn, oh ok [01:16] kgunn, ok you got silo 5 [01:16] sergiusens, you got silo 10 [01:19] thanks [01:19] tools always get 10 [01:19] hehe, luck of the draw! [01:21] robru: thank you sir [01:22] you are both welcome. sorry for the delay there I was afk [01:31] slangasek: hey, there's a merge conflict with your MRs; I'm guessing it's the changelog; do you want me to merge separately or in one stance? if you remove the changes from debian/changelog the train will generate one from the commit message set on the MR/MP [01:32] sergiusens: it is the changelog; happy to have the changelog entries removed to make the merge work; what's the best way forward? [01:32] sergiusens: they certainly don't need to be separate merges, so whatever's best to fix up the merge, let's do it [01:33] slangasek: just revert them [01:33] sergiusens: updated lp:~vorlon/phablet-tools/sane-adb-shells/ with changelog dropped - is that enough, or should I be dropping both changelog entries? [01:34] slangasek: drop both, or you'll only get that one entry [01:34] phooey ;) [01:34] sergiusens: done [01:35] thanks, building now [01:49] cihelp ping [01:49] cwayne_: that's not very helpful... [01:50] cjohnston, having some issues with the jenkins job http://s-jenkins.ubuntu-ci:8080/job/savilerow-trusty/25/console, seems its not actually pulling the latest rev somehow [01:53] as in, it's missing some files in the tarball that i know are in the branch.. [01:53] cwayne_: "revno:6" [01:54] cjohnston, right but it's missing the entire src/system/custom/lib dir [01:54] which is there if a do a fresh branch [01:57] cwayne_: seems like "tar Jcvf custom.tar.xz -C src/ system/" may be an issue [01:57] to me (not knowing anything about whats going on), it doesn't look like there would be a src/ and system/ [01:58] it's making the tarball itself fine though, it's just missing src/system/custom/lib [01:58] it's got src/system/custom/everything-else [01:59] cwayne_: is there a problem running it manually? [02:00] cjohnston, nope [02:01] just now ran it on a fresh branch [02:02] cjohnston, cwayne_, any chance someone did a push --overwrite on this? [02:02] fginther: it the branch seems to have it [02:02] + ls src/system/ [02:02] custom [02:03] ahh, I see what your saying.. [02:04] care if I retry a build after wiping the workspace? [02:04] +1 [02:04] === trainguard: IMAGE 271 building (started: 20140401 02:05) === [02:05] cwayne_, any harm in trying a clean rebuild? [02:05] fginther, none at all [02:07] nice, jenkins throws an exception trying to delete /home/ubuntu/jenkins/workspace/savilerow-trusty/src/system/custom/click/com.ubuntu.developer.alecu.qr-code/current. I'll check the host [02:12] cwayne_, the console looks like it found lib this time [02:12] fginther, awesome, thanks! [02:12] must've been a push --overwrite or something weird [02:13] cjohnston, cwayne_ 'current' was a symlink. I've seen a few jenkins bugs related to issues with symlinks [02:13] ah, could be [02:13] could we setup this job to always do a fresh branch? [02:15] seems like the job should only be firing when there is a code change [02:15] cwayne_, that might work, or the job may have to do some cleanup before it finishes [02:17] cwayne_, the job does a poll of the bzr tree, if the workspace is left in a bad state, the pollling may not go well [02:18] fginther, yeah, that makes sense [02:18] i can try and add some cleanup [02:19] thanks for the help fginther [02:22] and he leaves [02:22] cjohnston, /me thinks cwayne should be introduced to the jenkins charm [02:22] hehe [02:31] doanac: hey, care to run the andy job against https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AuDk72Lpx8U5dFlCc1VzeVZzWmdBZS11WERjdVc3dmc&usp=drive_web#gid=28 ? [03:34] === trainguard: IMAGE 271 DONE (finished: 20140401 03:35) === [03:34] === changelog: http://people.canonical.com/~ogra/touch-image-stats/271.changes === [04:12] morning [06:21] didrocks: bzoltan1 has set uitk to be tested but it'd need a packaging ack: http://162.213.34.102/job/landing-018-2-publish/lastSuccessfulBuild/artifact/packaging_changes_ubuntu-ui-toolkit_0.1.46+14.04.20140331-0ubuntu1.diff [06:22] Mirv: good to me, +1 :) [06:22] thanks [06:22] bzoltan1: re: your notes from other channel, this is your last chance to discuss if there's still some doubt on your mind :) [06:23] apparently sdk team has found a flakiness happening in clock app tests that shouldn't be coming from UITK, and sometimes all tests pass [06:23] but I guess the confusion was about why it's not flaky on the test infra or something [06:24] Mirv: correct [06:25] Mirv: the only flaky test is the ubuntu_clock_app.tests.test_stopwatch.TestStopwatch.test_click_clock_center_with_stopwatch_started_must_stop_it and the ubuntu_clock_app.tests.test_timer.TestTimer.test_delete_preset_must_remove_from_presets_list [06:26] Mirv: that means, that 2 out of 8 run failed ... [06:27] Mirv: http://paste.ubuntu.com/7187031/ http://paste.ubuntu.com/7187032/ [06:28] bzoltan1: ok. I guess we can assume at this point, especially as your people have studied it quite a lot, that it shoudn't be coming from UITK [06:28] bzoltan1: this test was never flaky beforehand, did you work with QA? [06:28] Mirv: I really do not think it has anything to do with the UITK [06:28] Mirv: we didn't have those testing failing over 20 runs [06:28] didrocks: Mirv: we worked with elopio [06:29] Mirv: so I think we should reallly have jibel's team expertise first [06:29] didrocks: I am not happy with that .. .but not landing the UITK will cause bigger issues soon [06:29] bzoltan1: what did elopio have to say? [06:29] bzoltan1: yeah, so then, you put the pain in the CI team to get the things fixed? [06:30] didrocks: Mirv: elopio said that he has been running clock tests on his mako with image 270 and the 18 silo ppa for one hour with no failures like 12 times [06:30] cihelp: seems that the tests don't unlock the screens? [06:30] didrocks: it'd be good to understand as much as possible. I know I can't replicate test infra results myself, for example yesterday notes-app failed for me consistently but ok on test infra. [06:30] didrocks: Mirv ^^ [06:30] bzoltan1: with your new UITK? [06:30] yes, silo 18 is the new UITK [06:30] with image 270 and the 18 silo ppa [06:30] bzoltan1: that sounds good at least [06:30] bzoltan1: ah, so this is good to me [06:31] Mirv: yes it does ... but I hate to see different test results than the dashboard shows [06:31] didrocks: I am confident, but I would like to see what the CI's dashboard shows [06:31] bzoltan1: so maybe a mental note to ask from elopio how he runs the tests, to try to get as good results in the future? [06:32] bzoltan1: that'll come with CI Airlines only unfortunately, getting the test infra to do the current manual work.. [06:32] Mirv: That is what I will do when elopio comes online [06:32] bzoltan1: like, if only you got it, someone else rerun it multiple times without getting it, knowing how hard and subtle flaky issues we can have, yeah, that's fine [06:32] plars: no email on the phone ML where #270 failed completely? [06:32] ok, but 12 times no failures from elopio is good enough, and I already know no other failures were found [06:33] so… I was telling… [06:33] bzoltan1: ah, so this is good to me [06:33] bzoltan1: like, if only you got it, someone else rerun it multiple times without getting it, knowing how hard and subtle flaky [06:33] issues we can have, yeah, that's fine [06:33] plars: no email on the phone ML where #270 failed completely? [06:33] did you get my messages? [06:34] depending on who you're asking from, yes I saw those three messages twice now [06:34] ok, I'm lagging/disconnected… [06:36] bzoltan1: published. it'll go to release team's unapproved queue, since it's also seeded on ubuntu/xubuntu/edubuntu desktop images [06:39] Mirv: Cool. Please ping me when I can merge to the trunk... [06:40] * bzoltan1 goes and starts to prepare the next landing :) [06:40] sure [07:01] Morning everyone [07:01] didrocks: so, in the end, we decided to drop the PHONECON ruleset as defined during vUDS? [07:05] Mirv: who should we ping from th release team to approve the UITK package? [07:05] good morning sil2100 [07:05] sil2100: seems so… [07:06] sil2100: also as we can't have any good test results for quite a while… [07:06] sil2100: do you feel better today? [07:06] didrocks: smoketest infra problems again? :( [07:07] didrocks: a bit, yes - I guess the cold and fever are slwoly giving up [07:11] morning sil2100! [07:12] bzoltan1: just hang around at #ubuntu-release and you can ask generally if you're in a hurry or if it seems it takes extra amount of time for them to review it. in general they go in the order of uploads, see https://launchpad.net/ubuntu/trusty/+queue?queue_state=1&queue_text= for the queue [07:18] sil2100: I don't know, we need to wait for the CI team [07:18] sil2100: however, the previous days, no crashers reported (due to the image…) [07:21] :| [07:30] sil2100, didrocks silo 20 is tested and good from my side. Would appreciate a round of exploratory testing [07:30] tvoss: sure, let me take a quick look [07:30] sil2100, thx [07:37] hm, strange, latest image doesn't boot for me, I wonder if I need to reflash or something [07:38] I mean, it boots but no unity8 can be seen [07:39] sil2100: I'm upgrading, will tell you in a few :) [07:40] * didrocks reboots [07:40] sil2100: anything in upstart logs? [07:42] sil2100: yeah, reproducing [07:42] (process:2351): GLib-GIO-ERROR **: Settings schema 'com.canonical.Unity.Launcher' is not installed [07:43] Yeah, I see the same, you think this can be the root cause? [07:43] it is [07:43] gsettings segfaults when the schema isn't available [07:43] Oh, how lovely [07:43] it's intended… :/ [07:43] ok, so the schema is shipped by /usr/share/glib-2.0/schemas/com.canonical.Unity.gschema.xml [07:44] so libunity-core-6.0-9 [07:45] "Drop unity-core based scopes in scopes plugin, expose attributes to shell" [07:45] I wonder if it's not what retained it into the image [07:46] and yeah [07:46] working then [07:47] plugins/Unity/Launcher/backend/launcherbackend.cpp: QGSettings gSettings("com.canonical.Unity.Launcher", "/com/canonical/unity/launcher/"); [07:47] so, it needs a dep on libunity-core [07:47] hm, why was this dropped from the image? [07:47] yep [07:48] Did we seed it then somehow? Since I don't see in the package diff any updates that could have dropped it [07:48] see my previous line [07:48] it was pulled by the scopes [07:48] not anymore [07:48] Ah, right, -plugin-scopes, damn, I'm blind [07:49] didrocks: should I poke Saviq to include the dep, or should we do it ourselves? A dep in unity8 I guess? [07:50] sil2100: no, I'm doing an upload [07:50] ACK [07:50] sil2100: didn't merge it back though, as I think the proper fix is to remove the schema from that package [07:50] have it on its own [07:50] and dep on that [07:52] Saviq: once you read that ^ [07:53] 7.84+14.04.20140327.1-0ubuntu2 uploaded [07:54] cihelp: unping on latest image not providing results. Would be good though to know about the previous one though [07:55] didrocks: ack, looking at it now [07:55] thanks psivaa :) [07:58] Let me drink some meds to be less coldish for the meeting [07:59] oki === psivaa changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: psivaa | CI Train support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but choose right timezone | Known issues: - [08:02] Saviq: let's talk about the best strategy, as we'll have click vs non click, maybe having unity8 shipping its own schema will make rather sense (as there is no more unity-common) [08:10] sil2100, didrocks, I don't think we should need it... [08:10] * Saviq greps [08:10] Saviq: you do in plugins/Unity/Launcher/backend/launcherbackend.cpp [08:11] ah right [08:11] we share the same launcher list [08:11] yep [08:11] I think with click, it doesn't make sense anymore, wdyt? [08:11] didrocks, you uploaded a fix directly, did you? and now we should fix in our trunk? [08:12] didrocks, we won't only support click [08:12] Saviq: yeah, but I think it's not really a fix (to depend on libunity-core for that) [08:12] didrocks, when we move to desktop [08:12] Saviq: yeah, but unity7 will never support click [08:12] so I guess having 2 list of launcher icon + handling the transition would make sense [08:12] didrocks, mhm [08:13] but now we'd need to handle the transition one way... and then the other... :| [08:13] didrocks: the apparent failures with 270 is due to a dashboard issue of mixing results from prev runs. trying to sort that out [08:13] didrocks, can we patch-aid it with the dep for now, and fix it properly through a bug? [08:13] Saviq: just one way I would say [08:13] Saviq: that works as well for me [08:13] psivaa: oh? [08:14] didrocks, it's not just us, we'd need to let customization know that the key changed and such [08:14] and put that key in a schema somewhere (ubuntu-touch-schemas?) [08:14] Saviq: oh right customization, let's do baby-steps then [08:14] Saviq: let me propose a branch then :) [08:14] didrocks: yea, http://ci.ubuntu.com/smokeng/trusty/touch/mako/270:20140331.1:20140331/7488/ubuntu_filemanager_app/ shows the passed ones and the failed prev run of the same image [08:14] Saviq: in few minutes :) [08:14] didrocks, thanks [08:14] * Saviq files a bug then [08:21] imgbot, status 271 [08:21] Image 271 test results on mako - Total: 134, Pass: 110, Crashes: 4, Rate: 80% [08:22] urgh [08:25] didrocks: todays image is a bit ummm broken for me [08:25] davmor2: yeah [08:25] davmor2: I uploaded a new unity8 [08:25] didrocks: admittedly I can only find one fault with it [08:25] didrocks: oh so it's all your fault :P [08:26] davmor2: no [08:26] davmor2: I pushed the fixed [08:26] we can maybe wait 4 minutes and discuss in the meeting? :p [08:27] ;) [08:27] sil2100: \o/ you feeling better dude? [08:30] at least all tests have 80% [08:30] thats something :P [08:33] ogra_: and I can't break the image at all it is 100% daveproof "Ship it!" ;) [08:33] haha [08:33] well xnox uploaded a change to adb [08:34] i know he ran AP tests with it but he most likely didnt test screen unlock [08:38] davmor2: more or less [09:01] popey: davmor2: is this the one, can I get some AP test report or such on it? http://s-jenkins.ubuntu-ci:8080/job/camera-app-click/48/artifact/out/com.ubuntu.camera_2.9.1.258_armhf.click [09:07] Mirv: ack [09:08] davmor2: are you guys saying unity doesn't start on #271 ? [09:09] popey: that is correct you just get a black screen [09:09] works here [09:09] phablet 2546 9.7 6.6 431676 125564 ? Ssl 10:06 0:17 unity8 [09:09] current build number: 271 [09:09] how odd. [09:09] didrocks: ^ [09:10] popey: you have better #271 than us [09:10] bet this is because my phone was rw when i updated [09:10] probably some lib that is still hanging around from #270? [09:11] Are the calendar-app AP tests known to be dodgy when run locally? [09:12] Laney: not that I'm aware, bugs welcome. [09:13] hmm [09:13] lemme see [09:18] ogra_: https://bugs.launchpad.net/ubuntu/+source/android/+bug/1287208 [09:18] Ubuntu bug 1287208 in android (Ubuntu) "ubuntu-device-flash doesn't remove the .writable_image flag" [Undecided,Fix committed] [09:18] davmor2, thanks [09:27] * didrocks reboots to previous kernel [09:31] Mirv: When you run the calendar-app tests using that recipe you gave me yesterday, does test_monthview_go_to_today_prev_year work for you? [09:39] tvoss: Laney: seb128: mhr3: Mirv: sil2100: going to stop CI train to move to prodstack in few minutes (of course, current builds aren't be stopped) [09:39] expect the downtime to be < 10 minutes [09:39] didrocks, ok [09:39] great [09:39] ok, good luck! [09:39] thanks :) [09:39] didrocks, good luck [09:40] will tell you when I shut down jenkins [09:40] thx for headsup [09:40] Laney, having fun with autopilot since yesterday it seems ... is all that only to test your qt patch from the other day? [09:40] ues [09:40] yes [09:40] velocity! [09:40] crazyness [09:40] you patched a "leaf" api [09:40] anyway the same thing happens with the old version of qtdeclarative [09:43] didrocks: ACK :) [09:43] so Not My Problem™ [09:43] Laney: yeah, just don't regress the image, only what we ask :) [09:43] don't fix stuff you can't/are not supposed to fix :) [09:44] it's just disturbing when things are different locally vs. in CI [09:44] Laney: yeah, agreed :/ [09:44] * seb128 impressed by Laney's perseverance [09:44] I wouldn't have bothered to run all autopilot tests for a 1 liner to a tz function [09:44] haha [09:58] Laney: it did succeed yesterday when I tested new UITK, ie calendar-app 12 tests passed (if I'm reading the correct line). but it's not like unheard that running AP tests locally does not easily yield same results as on the test infra [09:58] mmm [09:59] I guess the normal "pass" is "same result as without the update" when doing the local tests [09:59] Mirv: camera passes, stick it in the store and I'll get it approved [10:03] popey: ok [10:11] popey: it says package scan took too long, and that one should check status at https://myapps.developer.ubuntu.com/dev/api/click-scan-complete/93/ [10:12] Mirv: i cant see that ☹ [10:12] before that it did say it got submitted to https://myapps.developer.ubuntu.com/dev/api/click-upload/com.ubuntu.camera/ [10:12] or that [10:12] well it does say "Please check the status later at...", so I wonder who much later [10:12] we might need sergiusens then [10:13] yeah, i don't have the creds for that account, so can't see it [10:13] Morning all [10:14] Mirv: seb128: Saviq: sil2100: tvoss: mhr3: thostr_: stopping jenkins for the migration [10:14] didrocks: ok, good luck ;p [10:14] didrocks, have fun! [10:14] will try :p [10:15] ;) [10:16] didrocks, ok [10:16] popey: ok, I experimented what happens if I retry, and it worked.. Please check out the application at: https://myapps.developer.ubuntu.com/dev/click-apps/506/. [10:16] Mirv: i see it now [10:17] Mirv: approved it [10:17] nice! [10:25] all is transition, we have an issue on librarian firewall FYI [10:34] waiting on webops to check/open the firewall FYI [10:35] didrocks, seems there's a problem with SSO [10:35] didrocks, it redirected me to https://job/landing-002-1-build/build?delay=0sec [10:35] Saviq: please don't run anything for now [10:35] Saviq: sso was working for me [10:35] though [10:35] didrocks, sure, just saying [10:36] Saviq: let me unlog and relog [10:36] didrocks, it logged me in fine, but redirected to the wrong url [10:36] once, at least [10:37] didrocks: on 270 I was trying to rerun some things into late last night, and was hopeful that the previous failed attempts was just due to a typo in the test list [10:37] didrocks: but I fell asleep before it finished, so I didn't get to see until now, sorry [10:38] didrocks: I'm still not sure what happened there, but it looks like there is something wrong in the current image too [10:38] ev: I don't know why Saviq has that, seems that the juju charm was executed right? [10:38] didrocks: nothing changed in the ci code that could explain it though (no new unlocker or anything) [10:38] plars: ok, no worry, just an email on the ML next time :) [10:39] ev: may be the callback from sso side is wrong? [10:39] didrocks: sorry, intendted to [10:39] Saviq: weird, I'm redirected to the right url here [10:39] Saviq: let me try from another one [10:39] Saviq: yeah, working :/ [10:40] Saviq: and same from https://ci-train.ubuntu.com//job/landing-015-1-build/build?delay=0sec [10:40] Saviq: can you unlog/retry? [10:40] didrocks: that'd be surprising, given it redirected me just fine [10:40] didrocks, yeah let me try again [10:40] ev: yeah [10:41] I tried 3 configurations here [10:41] didrocks, ok, can't repro any more [10:41] yay [10:41] Saviq: can we blame you or your browser? :p [10:41] realistically, keep us posted if you see so [10:41] so… [10:42] it's one to watch out for though [10:42] stop using oxide :P [10:42] didrocks, you can [10:42] I doubt that came from nowhere [10:42] ev: agreed, let's see if more people get it [10:42] everything is back but the firewalling [10:42] so as long as you don't prepare or watch a build, you can resume working :) [10:42] (so publication + m&c) [10:48] 12:48:05 moon127 | didrocks, I've getting branch reviewed now. I would expect ~20 mins to land it on the firewall and make [10:48] | live. [10:48] Saviq: seb128: Mirv: sil2100: thostr_: mhr3: tvoss: ^ [10:49] Laney: ^ [10:49] (so, you can publish, m&c, just not build/watch) [10:49] neat [10:49] ok [10:50] ACK [10:50] 12:49:58 moon127 | didrocks, we've rushed that one through, and I can wget that file on the jenkins instance [10:51] so all should be back to normal, feel free to build things :) [10:51] Saviq: seb128: Mirv: sil2100: thostr_: mhr3: tvoss: Laney ^ [10:51] Yessss [10:51] ;) [10:51] yay for cloud [10:51] Thanks o/ [10:51] \o/ [10:51] lemme try to publish qtdeclarative [10:53] \o/ [10:53] didrocks: it seems I don't have permissions on the new jenkins to do builds! [10:53] Laney: some packages failed to build? [10:53] did they? [10:53] sil2100: did you check the ubuntu-unity box? [10:54] "saviq is missing the Job/Build permission" ;( [10:54] hm, right, it didn't even ask me that, just redirected instantly, let me relog [10:54] I believe spreadsheet is false false false [10:54] didrocks: it doesn't allow me to [10:54] sil2100, /me too, logging out and back in doesn't help [10:54] * Saviq goes to login.ubuntu.com [10:54] sil2100: also check to login from top right corner [10:54] didrocks: when I press 'login' then I just instantly get logged in without any prompts [10:55] hmm, yeah same here [10:55] Mirv: just did that [10:55] Saviq: sil2100: do you have the teams you are in checkbox? [10:55] didrocks: I got redirected to 'job' after SSOing [10:55] Laney: ^ [10:55] known? [10:55] didrocks: which checkbox? I have no checkboxes [10:55] is that what they are saying? [10:55] Laney: jobs? [10:55] Firefox can't find the server at job. [10:55] denied :| [10:55] ev: around? ^ [10:56] https://job/landing-008-1-build/build?delay=0sec [10:56] Laney: try again, I got the same the first time [10:56] Laney: hum, where are you clicking? [10:56] didrocks: when I press 'log in' I instantly get logged in, without any checkboxes [10:56] argh, so it's a real issue [10:56] didrocks, it's not asking for the team membership for some reason [10:56] build, SSO 2fa prompt, bogus redirect [10:56] didrocks: first time double clicking from spreadsheet it goes to https://job/ somehow after a couple of redirects [10:56] I didn't notice the checkboxes I'm afraid [10:56] the next time works [10:56] Mirv: I bet if you log out and back in it does it again [10:56] ev: really need you here ^ [10:57] anyway it works if you're already authenticated [10:57] just the error on first login [10:57] hi [10:57] (notwithstanding the permissions check that I didn't notice, sorry) [10:57] so first login is an issue? [10:57] didrocks, hi [10:57] hey moon127 [10:57] lemme try re-logging-in [10:57] so it seems first login is failing for everyone [10:57] and redirect to https://job/landing-008-1-build/build?delay=0sec for instance [10:58] didrocks: they shouldn't see the page with checkboxes as ci-train.u.c is in the SSO whitelist [10:58] ev: ah, ok [10:58] ev, didrocks: then how can I get my permissions correctly now? [10:58] * didrocks tries in a anonymous session [10:58] hmm I didn't get the 'job' thing that time [10:58] Mirv: are the permissions working then? [10:59] perhaps that only happens the first time you log in [10:59] ok, so here, on a anonymouse section, I login [10:59] then, I'm redirected [10:59] but unlogged [10:59] I reclick on "signing in" [10:59] ah no, I got it in a new browser that time [11:00] and I have rights [11:00] Laney: but then, do you have the right perms to run jobs? [11:00] didn't try yet [11:00] "timo-jyrinki is missing the Job/Build permission" [11:00] do you mind trying? (my user has more rights, so unsure it maps) [11:00] didrocks: ^ [11:00]