/srv/irclogs.ubuntu.com/2015/07/24/#ubuntu-ci-eng.txt

mardytrainguards: how can I prevent jenkins from generating a new changelog, and therefore avoid this error: https://ci-train.ubuntu.com/job/ubuntu-landing-001-1-build/206/console ?07:23
robrumardy: your packaging isn't really conforming to train expectations07:28
mardyrobru: apparmor-easyprof-ubuntu is not developed by us, I don't think I should change the packaging format07:29
robrumardy: then i guess you can't really release it in the train. Use a manual source package instead of mp.07:29
mardyrobru: or I guess I can modify the packaging format just for the train, and once the testing goes fine, we revert those packaging changes and land it as usual07:30
robrumardy: that sounds like a massive hassle, why bother? Just manually build a package and upload to the silo07:31
mardyrobru: ah, I didn't know that this could be done. Just by using dput, or how?07:31
robrumardy: yeah just make sure you're in the right team, or sil2100 can do it for you07:32
robrumardy: easiest might be to upload to a ppa you control, since i don't think you're in the team. sil2100 can copy it shortly (I'm literally in bed)07:34
abeatosil2100, mind assigning a silo for line 73?07:55
=== marcusto_ is now known as marcustomlinson
sil2100abeato: sure08:01
abeatosil2100, thks08:01
sil2100mardy: provide me with a package in a PPA or a source package and I'll upload :)08:01
sil2100mardy: https://wiki.ubuntu.com/citrain/LandingProcess#CI_Train_for_manual_source_uploads08:02
sil2100jibel, davmor2, popey: anything to discuss, or we can skip?08:03
jibelsil2100, nothing really important and the flow of incoming silos dropped08:08
davmor2sil2100: Loads to discuss....most about how cute your cat is....but not really work related not from last night anyway08:10
popeyhah08:10
sil2100The cat is indeed cute, just that my girlfriend gets tired of him biting her feet in the morning08:13
sil2100In the bed08:13
sil2100It's his way of saying: "I miss you mommy, get up and play with me"08:14
popeyours used to do that08:19
popeythey now get locked in the kitchen at night08:19
sil2100We usually throw him out of the bedroom when he does that too much and lock the door, still didn't learn how to use the door handle08:21
sil2100But I hope he'll just mature away from biting at night08:22
sil2100;)08:22
davmor2sil2100: does mean there is no meeting then08:25
sil2100davmor2: yep08:29
davmor2sil2100: you have to post pictures of your cat then :P08:29
popeyjibel: davmor2 got a new Reminders/Notes app for testing. Will anyone have some time to test today?08:46
sil2100Seeing the number of silos coming up, sure08:47
davmor2popey: possibly fill it in08:49
jibelpopey, sure, sil2100 is kindly proposing to do it himself :)08:51
sil2100...scratch that! Seeing the number of silos coming up, no way08:52
sil2100;)08:54
mandelsil2100, what does " is in the Proposed pocket" mean?08:54
mandelsil2100, also, does the packages from silo 09 sync with silo 17?08:55
sil2100mandel: it means the package is in wily-proposed08:55
mandelsil2100,  ok, great08:55
mandelsil2100, so, I'm wondering, the packages in silo 09, are they going to land in wily and then can they be sync to silo 17?08:55
sil2100mandel: if all is ok with the packages those should migrate to the release pocket soon08:55
mandelsil2100, sorry for the mess, trying to get this landed :-/08:55
mandelsil2100, ok, great08:56
sil2100Yeah, we can sync them then if needed :) Let's see what the excuses page tells us08:56
sil2100mandel: ok, I see an autopkgtest failed for u-d-m08:57
mandelsil2100, oh, can I see the error?08:57
sil2100Although it's hm, 'informational'08:57
sil2100Not sure what that means, maybe this autopkgtest doesn't block it08:57
mandelsil2100, where can I see that error? I should be able to judge :)08:58
mandelor maybe not hehe08:58
sil2100http://autopkgtest.ubuntu.com/packages/s/system-image/wily/amd64/ <- autopkgtest for system-image as ran because of u-d-m08:58
mandelsil2100, thx08:59
sil2100It's a cloud autopkgtest, first time I see those08:59
sil2100mandel: anyway, one autopkgtest is still running08:59
sil2100We'll see if it migrates once it's done08:59
mandelsil2100, I'm looking at the test anyway09:00
jibelsil2100, the cloud version is just informal and not blocking promotion IFAIK. The non-cloud version is there https://jenkins.qa.ubuntu.com/job/wily-adt-system-image/lastBuild/09:01
sil2100jibel: so as I expected, that's why I thought it probably doesn't block09:02
jibelinformational*09:02
sil2100Thanks for clearing it out ;)09:03
=== jamesh_ is now known as jamesh
jameshjibel: is there anything further that needs to happen before the thumbnailer can get out of proposed limbo?09:30
popeyjibel: added reminders/notes to the sheet, ready for testing thanks.09:32
sil2100popey: assigned in the sheet, QA should get it on the queue in a moment09:40
jibeljamesh, everything looks good on excuses.09:46
jibelinfinity, do you know why thumbnailer is a valid candidate and doesn't get out of proposed?09:46
sil2100jibel: looks like ubuntu-touch is uninstallable then... maybe the seeds need to be updated so that the removed library isn't pulled in?09:50
sil2100jibel: checking that now, but this might be the cause09:50
popeyta09:52
jameshsil2100: ubuntu-touch depends on thumbnailer-service, and there are deps on the QML plugin.  I don't see any dependencies on libthumbnailer0 (the removed package) outside of other binary packages provided by thumbnailer09:54
sil2100jamesh: same here, don't see anything in the dependencies, let me take a closer look09:56
dbarth__hey09:57
dbarth__for silo 31, i would like the silo to become wily only, as it's desktop oriented, and is really planned for 15.1009:58
sil2100dbarth__: ok, will convert it in a moment09:58
dbarth__super, thank you sil210009:58
dbarth__the silo has been tested, so we can then mark ok and land on wily only09:59
dbarth__sil2100: while we're at it, i think we could free silo 1, per robru's recommendation, and i will migrate to a ppa for the time being09:59
dbarth__and sort out the debian packaging there10:00
sil2100Yyyyh10:01
sil2100dbarth__: ok, makes sense, the earlier Yyyyh was to the issue I'm investigating now ;)10:01
sil2100jamesh, jibel: I think I have it10:01
sil2100eh10:03
sil2100I thought we fixed that with Mirv...10:04
sil2100jamesh, jibel: I'll try to unblock it in a moment, need to check the logs first10:05
sil2100Ah! I know what happened, we fixed that on vivid-overlay10:06
sil2100Didn't do the same fix for wily10:06
sil2100Ok, makes sense, fixing10:06
dbarth__can you help me get armhf builds enabled on that new ppa though: https://launchpad.net/~webapps/+archive/ubuntu/new10:07
dbarth__i'll move the branches over there10:07
dbarth__sil2100: ^^10:07
sil2100dbarth__: hey! I think we'd need someone like infinity and slangasek for that10:07
sil2100infinity, slangasek: ^10:07
=== chihchun is now known as chihchun_afk
sil2100dbarth__: silo 31 reconfigured, vivid package removed, please run a WATCH_ONLY build :)10:11
dbarth__sil2100: ok10:13
dbarth__https://ci-train.ubuntu.com/job/ubuntu-landing-031-1-build/13/console10:14
sil2100dbarth__: ok, it's ready now :)10:15
jameshsil2100: so what was holding it up?10:24
sil2100jamesh: qtdeclarative-opensource-src-gles... thumbnailer depends on latest ubuntu version symbols, so the -gles version needs to have the exact same version10:26
sil2100And it didn't, as the -gles and main version never had to follow the same versioning10:26
sil2100(mostly because the -gles packages are ubuntu-specific, where normal ones have debian revisions)10:26
sil2100So I had to bump qtdeclarative-opensource-src-gles from 5.4.2-0ubuntu3 to 5.4.2-1ubuntu310:27
sil2100So, in fact, we're now lying that this package had a debian releaase somewhere ;p But uh, no other way of resolving that without a lot of effort10:27
sil2100Building the package right now, I'll test it once it builds and publish10:28
anpokping trainguards10:57
anpokwe need an update of the packages in silo 05310:57
anpokwith the packages that landed from the mir 004 yesterday10:57
sil2100anpok: hm, did those migrate to the release pocket in the end? Did the boottest get whitelisted?11:00
anpoksil2100: yes, and fginther looked into improving the boottest to also enforce driver upgrades11:01
anpoksil2100: note.. silo004 contained a new mir package, than the one that got migrated from proposed to wily11:01
anpok*newer11:02
sil2100anpok: hmmm11:03
sil2100anpok: ok11:03
anpokyes big hmm11:03
anpoki made a fix attempt yesterday, that failed, reverted it...11:03
sil2100anpok: so basically you'd like to have all the packages that landed in the archive yesterday in silo 53, right?11:03
anpokall but gtk+3.16 but that didnt land since it was already outdated afaik. So, yes.11:04
sil2100gtk+3.16 doesn't make sense in the overlay anyway, so I'll remove it from the package-list11:04
anpoki think someone else already made a gtk+3.14 in vivid+o anyhow11:05
anpokso yes I think it is not necessary11:05
sil2100I'll reconfigure the silo to be a sync silo, and those packages that cannot be synced I'll manually upload11:05
sil2100Give me a few minutes to finish up what I was doing tho ;)11:05
anpokthank you!11:06
sil2100jamesh: publishing the fix, we should be unblocked soon11:08
jameshsil2100: yay!  Thanks for your help11:08
sil2100yw!11:08
sil2100anpok: on your request now11:10
sil2100uh, symbols file removal11:14
sil2100I can't publish that sadly11:15
sil2100Not without some other automated way of checking for abi-compliance11:15
sil2100jamesh, marcustomlinson: are you guys using something like dh-acc or abi-compliance-checker in an automated way in unity-scopes-api ?11:16
LaneySymbols files are also used to give reverse dependencies the right versions11:18
marcustomlinsonsil2100: no we're not. tvoss ^, any suggestions?11:18
LaneyThose other thing don't fully replace them11:18
sil2100Laney: indeed, you're right11:19
sil2100anpok: hey, is there a specific build-order required in the mir silo?11:19
marcustomlinsonsil2100: ok, putting symbols files back sadly11:29
dokosil2100, you really need to update qtdeclarative-opensource-src-gles - 5.4.2-0ubuntu5 (Newer version available) in silo 1611:38
sil2100doko: ok, I'll update the one in the PPA, since the package I pushed to the archive is just a version bump11:40
sil2100So I'll do a similar bump in the transition PPA11:40
sil2100doko: thanks for mentioning11:40
dokosil2100, sure, use the next version number then11:42
sil2100doko: pushed corrected, should be ok once it builds11:46
* sil2100 goes off to prepare lunch11:57
=== alan_g is now known as alan_g|lunch
=== _salem is now known as salem_
anpoksil2100: well we usually first have to build mir then the rest.. then -gles sync packages. But do you need that info for a sync silo?12:33
jdstrandmardy: how we do it is in the spreadsheet, add apparmor-easyprof-ubuntu to the 'Additional source packages to land' column, then dput directly to the silo ppa or pocket copy it into the silo ppa from another ppa12:35
jdstrandmardy: you can still do MPs for other stuff12:36
jdstrandmardy: in the same silo12:36
mardyjdstrand: OK, thanks, but after talking with dbarth we convened that it's better to use a PPA, not to block one silo for long (we might be testing this for quite some time)12:36
jdstrandmardy: ok. let me know if you need help with a silo at some future date12:37
dokosil2100, please could you fix the silo 16 and not leave it in a broken state?12:45
mandelsil2100, what is going on with silo 09? Just trying to understand :-/12:49
=== alan_g|lunch is now known as alan_g
anpoksil2100: problems in silo 53?13:34
sil2100anpok: hm, just got from lunch, let me take a look at that - it looks as if the upload got rejected13:36
anpokok, also the merge of the mps from silo004 hasnt happened yet (afaik) - or will that happen later and separately13:38
anpoksil2100: I am off for about two hours now.. will be back..13:38
=== marcusto_ is now known as marcustomlinson
seb128did anyone else notice bug #1478022 ?14:17
ubot5bug 1478022 in unity8 (Ubuntu) "Click previews slow to load and can't scroll while loading is not done" [Undecided,New] https://launchpad.net/bugs/147802214:17
seb128click details take like 10 seconds before to render any content here14:18
jibelseb128, it takes around 5s here. Is it specific to an app or you tried different ones?14:21
seb128jibel, I tried different ones14:21
seb128calendar, dekko14:21
jibelI tried the same14:21
jibelscrolling is really broken when it's loading though14:22
jibelwhen scrolling is slow dash uses 100% cpu14:23
seb128jibel, I reported https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1478022 as well14:25
ubot5Ubuntu bug 1478022 in unity8 (Ubuntu) "Click previews slow to load and can't scroll while loading is not done" [Undecided,New]14:25
seb128and https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/147802614:25
ubot5Ubuntu bug 1478026 in unity8 (Ubuntu) "Never loads click preview again if first tried offline" [Undecided,New]14:25
seb128those feel like regressions14:26
seb128I wonder if that's the same class of issues than bug #147628214:26
ubot5bug 1476282 in Today Scope "Today scope needs attention: slow to load and missing items" [Critical,Confirmed] https://launchpad.net/bugs/147628214:26
seb128pmcgowan, ^ unsure if others reported issues around those lines? we should also probably have that today's bug on the ota6 list, seems quite some users hit the "missing day info" issue14:27
abeatosil2100, could you publish silo 8 ? I am not sure whether I can do that or not14:29
sil2100abeato: on it, I have a meeting right now tho'14:30
abeatosil2100, nw, not in a hurry14:30
pmcgowanseb128, yes kyle has been trying to reproduce it etc14:31
seb128pmcgowan, I get it pretty easily, but I've other issues with other scopes cf the click preview ones I just listed, I wonder if there are some issues in the scope framework recent updates14:34
sil2100abeato: ^ could you approve some of the merges?14:39
abeatosil2100, working on that14:40
bfillersil2100: can we get Elleo (michael-sheldon) added for permissions for the ci-train spreadsheet, etc so he can create silos and build?15:00
sil2100bfiller: sure thing! I thought he had permissions already actually15:01
sil2100Let me add him15:01
bfillersil2100: oh, maybe he does15:01
sil2100bfiller: hm, he wasn't, I was sure he was... anyway, added now15:02
sil2100kgunn: I'm re-assigning silo 53 to another silo15:23
sil2100Since we seem to have some issues with source synces15:24
kgunn@vogons ^15:26
abeatosil2100, could you try now publishing silo 8 ?15:45
Laneytoo late15:47
LaneyI have all the power so did it15:47
abeato:)15:47
infinityjibel: What triggers the DKMS jenkins jobs?16:27
* jibel tries to remember the details16:29
davmor2infinity: the words jibel is looking for is Magic16:30
jibelinfinity, there a a daily job that verifies if there is a new dkms module or kernel in one of the target release/pocket or ppa16:30
infinityjibel: Okay, so I just need to be patient for that cronjob?16:30
infinityjibel: I tried triggering one by hand from the jenkins UI, and that seemed to do nothing. :P16:31
infinityOh, but maybe that was because it was wily-release, and it's outsmarting me.16:31
infinityBut there is no wily-proposed job, so I guess I wait.16:31
jibelinfinity, I noticed that the node they were running on died 10h ago, cihelp restarted it, but these jobs all failed and won't be restarted automatically16:31
jibelinfinity, if should do something :)16:32
jibelit*16:32
jibelinfinity, which one was it?16:32
infinityjibel: Well, this is the one I retried: http://d-jenkins.ubuntu-ci:8080/job/dkms-wily-release_canonical_kernel_team_ppa-generic-fglrx_core/16:32
infinityjibel: But the dkms package I'm wanting to verify fixed is in -proposed, not release, maybe it checks for versions changes in the release pocket and bails out of nothing's new?16:32
jibelinteresting: /tmp/hudson2415302381661435058.sh: 11: cd: can't cd to /home/auto-package-testing/dkms-testing/16:33
jibelinfinity, in this case ^ it a real problem16:33
infinityOh, fun.  Then consider the problem reported!16:34
infinityjibel: HEY JIBEL.  THERE'S A PROBLEM.16:34
* jibel -> EOW16:34
infinityHahaha.16:34
* infinity mails cheese and wine.16:34
infinityThat's what fuels French people, right?16:34
* infinity throws in a baguette for good measure.16:34
mandeltrainguards how do I bump the version number from 0.9 to 1? Is there some magic to make the bot do it?16:39
sil2100mandel: hey!16:39
mandelsil2100, though you were done for the day :)16:39
mandelsil2100, so the so is 1 gut the ppa creates 0.9-foo-bar16:40
sil2100mandel: just modify the debian/changelog of the package with a merge request in the train and add a new changelog entry with the new upstream version16:40
sil2100So, an example:16:40
mandeldch -i and the Bump version number due to blah blah :wq16:40
mandelsil2100, ^16:40
sil2100Let's say you have merges for project foo and the last changelog entry is 0.9+15.04.20150707-0ubuntu116:41
sil2100And you want to change to 1.016:41
sil2100THen you do dch -i, change the latest version number to 1.0 (or 1.0-0ubuntu1, doesn't matter) and get that through the train16:41
sil2100The CI Train will parse that and convert it to a CI Train number itself16:41
sil2100Just be sure to leave it as UNRELEASED16:42
mandelack16:42
slangaseksil2100: do you know if dbarth is specifically looking for https://launchpad.net/~webapps/+archive/ubuntu/new to be a devirt ppa, or just have armhf enabled?16:53
sil2100slangasek: hm, not sure, I think he might want to poke us later to perform binary syncs after testing is done, but I suppose we can do source-copies instead before release16:56
sil2100We'd need him to get to know his rationale16:56
slangaseksil2100: why would he be asking for binary syncs from here instead of using the silos?16:56
infinitysil2100: If there was to be binary copying, it would need to be devirt, with all arches on, and the right configs to match.  And Canonical-only, etc.16:57
slangasek(in other words: I don't think we would want to allow that)16:57
infinitysil2100: But my impression from scrollback was just that he wanted somewhere to play without having to lock a silo.16:57
sil2100infinity: might be16:57
infinityAnd "play" doesn't imply "release versions".16:57
sil2100As I said, we can do source-copies to silo PPAs before release16:57
slangasekthe main difference between a devirt and non-devirt ppa with armhf support is that non-devirt is emulated, and the emulation is sometimes buggy.16:57
infinityAnd s-l-o-w.16:58
=== alan_g is now known as alan_g|EOW
sil2100popey: did you publish the new notes app?17:08
popeysil2100: not yet, will do shortly17:09
sil2100popey: ok, thanks :)17:09
davmor2sil2100: it only just passed17:11
popey:)17:12
=== balloons_ is now known as balloons
=== balloons39 is now known as balloons
popeysil2100: done17:20
sil2100popey: thanks! Yay :)17:20
kenvandinemandel, can silo 9 be published again?18:22
anpoksil2100: do you know when ci will merge the MPs of silo004?18:31
kgunntrainguards you can blow away silo 42, that's old/stale...do i just delete the cooresponding row in the sheet ??18:37
kgunndidn't want to inadvertently break somthing18:37
robrukgunn: yep you can delete rows no problem.18:37
kgunnanpok: it needs testing set to passed i think18:39
kgunnit says "no" in the sheet right now18:39
robruanpok: they get merged after they get published. silo says 'packages built'18:39
anpokhum18:40
anpokthat info is gone again18:41
kgunnnow they're ready18:43
kgunnanpok: are you going to have a beer or 2 this weekend after all this :D18:43
anpokkgunn: it isnt over yet!!!18:43
kgunnlol18:43
kgunntrue18:43
kgunnanpok: if those land and migrate ok, i can start testing for vivid+o this afternoon....18:44
robruanpok: do you know anything about this gtk upload that happened without you? https://launchpad.net/ubuntu/+source/gtk+3.0/3.16.6-1ubuntu118:44
anpokrobru: yeah. there was a problem in the gtk+3.16 package .. and I believe seb fixed it and uploaded it independent from the silo18:45
robruanpok: ok I'll remove it from the silo then since we can't publish a lower version over the higher one in proposed18:45
anpokyes18:45
anpokrobru: i would like to amend that mir0.14.0 was already released to wily18:47
robruanpok: ok?18:47
anpoknot sure what happened to the row state18:47
robruanpok: the silo has packages that are newly build and not released yet to wily right?18:47
anpokes18:47
anpokyes18:47
kyrofatrainguards: Does the CI train support launchpad git the same as bzr?19:10
robrukyrofa: nope, not at all19:16
kyrofarobru, thanks for the info. Do you know if that's on the roadmap at all?19:19
robrukyrofa: in theory I suppose. lots of other priorities right now though. don't hold your breath, sorry19:20
kyrofarobru, alright, thanks!19:20
robrukyrofa: you're welcome19:24
kgunnanpok:  so do you need to "build" something ? seems like the gtk upload changed the status once again19:44
kgunn"ready to build"19:44
anpokkgunn: no clue whats going on19:52
kgunnrobru: does anpok just need to do a rebuild of 1 pkg or something ? ^19:53
robrukgunn: anpok oh sorry guys, I was fiddling around with that and then got distracted19:55
robrukgunn: anpok: I had to reconfigure to remove gtk because the version in the silo is lower than the version in archive. will publish shortly19:55
mandelkenvandine, yes, sorry had to bump the version number19:57
kenvandinemandel, i pushed the button :)19:58
mandelkenvandine, no need to recompile the other packages, it was a simple chalog change19:58
mandelkenvandine, oh, hehe19:58
mandelkenvandine, sorry I had dinner (22:00) :)19:58
kenvandineno worries man19:58
kenvandinei figured you were gone so took care of it19:58
mandelkenvandine, awesome, thx19:59
kenvandinenp19:59
robrukgunn: anpok: just want to confirm with you guys that the only packages that changed since last publisher are xorg and mir?20:06
anpokxorg changed?20:06
kgunnanpok: robru that wouldn't be completely shocking...silo 4 has been around 4 weeks, robert ancell and co prolly took in a new version of xorg20:09
robrukgunn: anpok: the new xorg is from the 15th, and appears not to be in wily yet.20:12
robrukgunn: anpok: ok for some reason the train is still failing to recognize that glmark is in the archive already, so if you happen to notice mir and xorg migrate (eg, the status will only say "glmark is in the silo" and won't mention any other packages), ping me and I can make the merge happen.20:13
robrujamesh: does silo 30 not need qa?20:34
kgunnrobru: yo, silo stuck on glmark2 now20:55
=== salem_ is now known as _salem
robrukgunn: anpok: https://ci-train.ubuntu.com/job/ubuntu-landing-004-3-merge-clean/27/console lol, 16 builds of mir in total21:49
=== _salem is now known as salem_
=== salem_ is now known as _salem
=== ev_ is now known as ev

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!