/srv/irclogs.ubuntu.com/2016/05/09/#ubuntu-ci-eng.txt

=== chihchun_afk is now known as chihchun
Mirvsil2100: too much for you? :)08:33
=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
pstolowskijibel, hello! quick question, the yellow 'next milestone' tag you have on a few silos in the dashboard means they didn't make it for OTA11?11:07
jibelpstolowski, yes, unless it something critical for OTA11 of course. which silo?11:10
jibelit is*11:10
pstolowskijibel, 4711:10
pstolowskijibel, nothing super critical, a couple of nice-to-have bugfixes. leave it up to you to judge the importance, i understand you're under pressure already11:12
jibelpstolowski, it will unlikely make it, there are still too many things to land today and testing should have started yesterday.11:14
pstolowskijibel, i see, ack. thanks!11:16
jibelpstolowski, besides a fix for bug 1578283 is a good candidate for OTA1111:17
ubot5bug 1578283 in Canonical System Image "App scope is blank" [Critical,Confirmed] https://launchpad.net/bugs/157828311:17
pstolowskijibel, i know, but atm we still don't know what's causing it and can't reproduce on our devices11:18
jibelpstolowski, I'll try to reproduce with the upstart job method, I saw it several times11:19
jibelpstolowski, like right now on arale after a reboot11:20
jibel(but without debug mode on)11:20
pstolowskimarcustomlinson, ^11:25
marcustomlinsonpstolowski: yeah I'm really stumped with this one11:28
pstolowskijibel, do you recall when did it start happening (roughly)?11:28
jibelpstolowski, https://bugs.launchpad.net/ubuntu/+source/unity-scope-click/+bug/1578283/+attachment/4659168/+files/scope-registry.log11:35
ubot5Ubuntu bug 1578283 in Canonical System Image "App scope is blank" [Critical,Confirmed]11:35
jibelpstolowski, I don't remember sorry11:35
jibelmarcustomlinson, ^11:35
pstolowskijibel, thanks11:36
jibelpstolowski, marcustomlinson All I did is reboot, unlock the SIM then unlock the greeter. the progress bar at the bottom of the app scope is moving a moment then stops and the scope remains blank11:37
pstolowskijibel, can you also attach unity8-dash.log?11:37
jibelpstolowski, attached11:39
jibelthese log files really need timestamps11:39
pstolowskiindeed11:41
rvrChrisTownsend: bregma: 15 and 56 approved12:11
ChrisTownsendrvr: Sweet!  Thanks!12:11
rvrChrisTownsend: However, I didn't get the gedit icon on the Launcher12:11
rvrNot even with the new puritine click12:11
ChrisTownsendrvr: Really?  Did you restart the frieza?  I had to reboot it to get it to work.12:12
rvrYes, I rebooted a couple of times12:12
rvrIt doesn't appear in the Libertine scope either (checked 15+56 together).12:12
ChrisTownsendrvr: Hmm, I'm not really sure why you don't see it then.12:13
ChrisTownsendrvr: Just to make sure, click list shows version 0.8 of puritine installed?12:14
rvrChrisTownsend: Yes, see the trello card12:27
rvr$ click list | grep puritine12:27
rvrcom.ubuntu.puritine 0.812:27
ChrisTownsendrvr: Ok, well, I'm really at a loss as to what is going on.  I see it on my frieza (only after a reboot).  I'll keep on eye on that and thanks for letting me know.12:29
rvrChrisTownsend: Which image are you using?12:30
ChrisTownsendrvr: Let me check12:30
ChrisTownsendrvr: As another way to see if the icon theme is properly installed in the puritine click, whenever you do open Gedit, do you see proper icons in the gedit menu bar?12:31
rvrChrisTownsend: Nope12:32
rvrChrisTownsend: Maybe it needs a new gedit click ?12:32
Saviqplars, hey, I'm afraid krillin-07 is dead on us, lifeboat can't recover it since a few days12:33
Saviqarale still unstable on the USB connection, too :(12:34
ChrisTownsendrvr: Well, puritine is all one big click.  That version 0.8 should have the humanity-icon-theme in it.  I wonder if somehow the wrong puritine click was downloaded.  I created a different 0.8 that didn't have the theme in it, and I'm wondering if somehow that was left laying around.12:34
ChrisTownsendrvr: BTW, I'm using build 97 form rc-proposed.12:34
ChrisTownsend*from12:34
rvrMe too12:34
ChrisTownsendrvr: Do you have time to try downloading the Puritine click again and reinstall it on your frieza?12:35
ChrisTownsendrvr: I'll give you a different link.12:35
Saviqjibel, FYI, https://requests.ci-train.ubuntu.com/#/ticket/1381 is ready (just waiting for britney results, but our own were happy - and we've fixed one of the flaky tests we've recently seen)12:36
rvrChrisTownsend: Yes12:36
ChrisTownsendrvr: Ok, cool.  Try this one: https://private-fileshare.canonical.com/~townsend/puritine/com.ubuntu.puritine_0.8_armhf.click12:37
ChrisTownsendrvr: This is the one I have installed on my frieza and gedit icons work.  *Hopefully* it will work for you.12:37
Saviqjibel, sorry we've missed the mark on Final Freeze12:38
jibelSaviq, it is really too late to land it for OTA1112:39
jibelthere are still silos in the queue and several critical bugs to fix12:39
rvrChrisTownsend: Downloading12:40
Saviqjibel, you guys' call - these are all really small fixes for DGU and test stability - if we can't land for OTA11, bummer - maybe we cherry pick some of them that you decide worth it12:41
jibelSaviq, does it fix anything in this list https://bugs.launchpad.net/canonical-devices-system-image/+bugs?field.tag=lt-blocker ?12:41
Saviqjibel, no12:42
Saviqjibel, I'd venture to say a few of the ones it fixes, should be added to that list, though ;) - but again, your call12:43
kdubdo the launchpad ppa's use pbuilder? trying to debug an issues that only happens in the builder12:46
Saviqkdub, sbuild12:47
kdubhmm, makes my problem more mysterious, as local sbuild works for me12:47
kdubthanks for confirming Saviq12:47
ChrisTownsendsil2100: Hey!  Are you able to ack the packaging changes in https://requests.ci-train.ubuntu.com/#/ticket/1360 and publish?  If so, do you have time to do that?12:52
sil2100ChrisTownsend: hey! Yes, I have the power, but would need a few minutes12:53
ChrisTownsendsil2100: No worries if you don't:)12:53
ChrisTownsendsil2100: Ok, thanks.  A few minutes is fine.  I know you're busy though, so I can find someone else if you need me to.12:54
plarsSaviq: thanks for letting me know. I'll look into it in just a bit13:00
rvrChrisTownsend: Installed the new click, same result13:09
rvrcom.ubuntu.puritine0.813:10
ChrisTownsendrvr: Ok, could you do a ls -la on ~/.cache/libertine-container/puritine?13:10
rvrChrisTownsend: http://paste.ubuntu.com/16317818/13:11
ChrisTownsendrvr: Ah, there's the problem.13:11
ChrisTownsendrvr: So remove ~/.cache/libertine-container/puritine and then do13:12
ChrisTownsendrvr: initctl --session start purtine-click13:12
ChrisTownsendrvr: Then send me the ls -la output again,13:12
rvrinitctl: Unknown job: purtine-click13:13
* ChrisTownsend scratches head13:13
rvrAh, wrong spelling13:13
rvrDone13:14
ChrisTownsendrvr: lol, I misspelled it13:14
rvrlrwxrwxrwx 1 phablet phablet  110 may  9 14:13 rootfs -> /opt/click.ubuntu.com/.click/users/@all/com.ubuntu.puritine/libertine-data/libertine-container/puritine/rootfs13:14
ChrisTownsendrvr: Ok, that is better.  Try it again.13:14
ChrisTownsendrvr: BTW, this won't be a problem on the channel image since penk will be installing the updated click over the old one in /custom13:16
ChrisTownsendrvr: Another thing, the puritine-click upstart job never took into consideration that the puritine click could be installed in 2 separate places.  In the next Libertine release, I will have working puritine click hook that will replace the upstart job and can handle multiple puritine-type click packages as well as a puritine click being installed in 2 separate places.13:19
rvrChrisTownsend: Cool. This looks way better.13:20
ChrisTownsendrvr: Awesome!13:21
rvrI can see the icon in the Launcher, in the scope, and gedit displays nice icons.13:21
rvrWee!13:21
rvrChrisTownsend: So, what will happen to people with old symbolic links like me?13:22
ChrisTownsendrvr: Well, when the next frieza image is released, it will just work since penk rolls everything up in the custom tarball.13:24
ChrisTownsendrvr: If someone installed the puritine click via pkcon, then it will still work since puritine is in it's standard place.13:25
rvrOk13:25
ChrisTownsendrvr: Thanks for verifying this.13:26
rvrYou're welcome13:26
=== _salem is now known as salem_
tedgHuh, apparently I can't merge a silo? I thought any lander could do that.14:16
tedgsil2100: Can you merge this for me? https://requests.ci-train.ubuntu.com/#/ticket/136014:59
sil2100tedg: it looks landed to me - what's wrong with it?15:00
tedgsil2100: Oh, I thought I needed to merge at the end...15:02
tedgsil2100: Seems that happened automagically. Sorry.15:02
sil2100tedg: no no, that happens automatically usually, only in some specific cases when the package is blocked etc.15:03
marcustomlinsonfaenil, jibel, pmcgowan, popey: Hey guys, could you please install the click scope from this PPA: https://code.launchpad.net/~unity-api-team/+archive/ubuntu/dev-build-1, and let me know if you still get an empty click scope.15:39
faenilmarcustomlinson: ok15:40
kenvandinesil2100, any idea when we'll get a new turbo image?  we just discovered that bug 1579626 is already fixed in the archive but not in the turbo image15:45
ubot5bug 1579626 in messaging-app (Ubuntu) "Cannot send SMS with single SIM card" [Critical,Confirmed] https://launchpad.net/bugs/157962615:45
pmcgowanmarcustomlinson, all 3 packages there?15:49
marcustomlinsonpmcgowan: yeah, should be unity-scope-click-departmentsdb, unity-scope-click-init-departments, and unity-scope-clic15:50
pmcgowanack15:52
popeymarcustomlinson: ok15:59
popeymarcustomlinson: is there some magic to install that ppa, or just grab the deb and dpkg -i it, enough?16:00
marcustomlinsonpopey: that is enough yeah. Ensure you install unity-scope-click-departmentsdb, unity-scope-click-init-departments, and unity-scope-click16:00
popeyok16:01
rvrrenatu: Is there any way to trigger bug 1577416?16:04
ubot5bug 1577416 in Canonical System Image ""error code from SyncEvolution fatal error (local, status 10500): no datastores active, check configuration"" [High,In progress] https://launchpad.net/bugs/157741616:04
=== chihchun is now known as chihchun_afk
popeymarcustomlinson: hard to say this solved it, given it doesn't always happen for me16:17
marcustomlinsonpopey: yeah this is a tricky one. I'm yet to see it at all myself16:18
faenilmarcustomlinson: I don't see departmentsdb here https://code.launchpad.net/~unity-api-team/+archive/ubuntu/dev-build-1/+build/971369116:20
popeyfaenil: https://code.launchpad.net/~unity-api-team/+archive/ubuntu/dev-build-1/+files/unity-scope-click-departmentsdb_0.1.1+16.04.20160415-0~444~ubuntu15.04.1_all.deb16:21
marcustomlinsonfaenil: yeah it ends up under the amd64 build (_all.deb): https://code.launchpad.net/~unity-api-team/+archive/ubuntu/dev-build-1/+build/971369016:21
faenilah ok, right16:21
faenilpopey: cheers16:22
faenilmarcustomlinson: at least after 1st reboot no problem, scope was populated16:27
popeyfaenil: marcustomlinson yeah, i rebooted a bunch of times and couldn't reproduce it. But typically for me it happens when on slow/rubbish networks16:36
popeyTry it in the canonical office ㋛16:36
marcustomlinsonpopey: interesting. let me try edge16:36
popeyCould be a coincidence of course16:37
* faenil is in the office :)16:37
rvrrenatu: re-ping17:33
marcustomlinsonfaenil, jibel, pmcgowan, popey: I'm gonna infer by the inability to reproduce an empty apps scope, that it might actually be resolved? Please let me know if you see it again. +This fix (hopefully) is ready in silo 54 now.17:59
pmcgowanmarcustomlinson, just installed the PPA and doing a million reboots18:00
pmcgowanwhat did you change?18:00
marcustomlinsonpmcgowan: there was only one significant change to the click scope since March: https://code.launchpad.net/~dobey/unity-scope-click/really-we-want-to-sign-all-the-requests/+merge/29076318:01
pmcgowanoy did we revert that again?18:01
marcustomlinsonpmcgowan: yeah18:01
marcustomlinsonIt's a rocky one18:01
pmcgowanmarcustomlinson, so I occasionally (one in 10) see a black scope then the icon then finally the scope contents18:02
pmcgowanso some other race as well18:02
marcustomlinsonif you unlock quick enough I guess you'll catch the black screen before unity8-dash even starts18:03
robruslangasek: meeting?18:03
pmcgowanmarcustomlinson, yeah but sometimes it takes like 10 secs vs no time18:03
marcustomlinsonhmm18:03
marcustomlinsonok fair point.18:03
pmcgowanso it just worked but the contents filled in late, as opposed to the silo 71 fix to prepoulate18:04
pmcgowanits different every boot for me18:04
pmcgowanmarcustomlinson, you didnt revert the prepoulate change did you?18:05
marcustomlinsonpmcgowan: If you were on the lock screen for that 10s, then unlocked say 5s later, you'd have probably seen the scope populated18:05
marcustomlinsonpmcgowan: no I didn'18:05
marcustomlinsondidn't18:05
pmcgowanok let me but more patient18:05
pmcgowanmarcustomlinson, hmm still no18:06
pmcgowanblack18:06
pmcgowanwaited 10+secs18:06
marcustomlinsonpmcgowan: that is really odd, the prepopulate is a unity-scopes-shell fix18:07
pmcgowannot saying its related18:07
marcustomlinsonthe only package I changed was the click scope18:07
marcustomlinsonah ok18:07
pmcgowanI occasionally saw it wihtout it18:07
pmcgowanjust rebooting a lot right now :)18:07
marcustomlinsonha18:07
pmcgowanhmm18:07
pmcgowanmost ervy time18:08
renaturvr, hi18:11
rvrrenatu: Hi18:12
rvrrenatu: I am leaving now18:12
rvrrenatu: I was waiting for feedback for silos 71 and 3618:12
renaturvr, any problem with the silo 36?18:13
rvrrenatu: I checked silo 71 and sync'd ok18:13
rvrrenatu: ToyKeeper said "Did the fix for lp:1434576 (quick scroll) get left out? I don't see anything attached which looks like a fix for that, and I can still reproduce the issue with the silo installed."18:13
rvrBut I really have to leave now18:13
rvrrenatu: On 71, I was asking for a way to reproduce the fix, or check that it was fixed18:14
rvrPlease, comment and I'll check when I return18:15
renatuok18:15
renatu71, there is no reliable way to reproduce the error18:15
plarsSaviq: krillin-07 should be back now19:05
Saviqplars, thank you19:05
plarsSaviq: happy to help19:06
Saviqplars, I'm afraid to say arale-01 does not have a good track record https://unity8-jenkins.ubuntu.com/computer/arale-01/builds :/19:13
Saviqall the red ones are some sort of connection issues afaict19:14
Saviqkrillin, when it worked, was in a much better state https://unity8-jenkins.ubuntu.com/computer/krillin-07/builds (the most recent red ones were due to its recent failure, most of the ones before are actually "real" failures due to image issues)19:15
plarsSaviq: same problem that you had with arale-03?19:17
plarsSaviq: or just not recovering correctly?19:17
Saviqplars, it's difficult to say, the errors as surfaced are of all kinds, happening at random times - sometimes it won't flash, sometimes it will die mid-test19:17
Saviqplars, I might need to bring back some debugging into those jobs to see what's happening (tail -n100 syslog would maybe tell us if they are indeed usb issues)19:18
plarsSaviq: I'm trying to recover arale-01 now, an it seems to be recovering at least19:25
Saviqplars, looking at the recent failed runs, none of them successfully flashed anything - be it during ./recover or later u-d-f19:28
plarsSaviq: it seems to be in the middle of flashing a stable image with recover right now, I'll let you know how it goes19:29
Saviqplars, tx19:29
plarsSaviq: arale-01 recovered just fine for me, but these aren't the only devices exhibiting weird behavior today. Rick will be back in the datacenter tomorrow and may be able to do some more hands-on investigation.19:47
Saviqplars, ack, let me run something on it19:55
tedgI thought if I needed a lib version that is in xenial in the vivid overlay I added it to source packages. Is that not correct? It didn't seem to grab it.20:24
Saviqtedg, you need to copy it, too20:28
Saviqtedg, trainguards and core devs have the rights to do it20:28
tedgSaviq: Ah, I didn't think I had a button for that :-)20:28
tedgPerhaps kenvandine can help me?20:29
SaviqI vaguely remember there was something like a crazy copy:source:destination syntax, not sure where that was though and/or if it's still available20:29
kenvandinetedg, you need a xenial package copied to a landing ppa?20:30
tedgkenvandine: Yes, whoopsie to landing 57 please20:30
kenvandinetedg, done20:33
tedgCool, thanks kenvandine !20:33
kenvandinenp20:33
robruSaviq: that never worked for this purpose. That was for copying from Ubuntu to ubuntu-rtm. Disabled for now because it stopped working and nobody noticed20:33
tedgUhg, pyflakes3 isn't in Vivid :-(20:38
robrutedg: fun times. IIRC i once backported it to trusty and it built without needing to backport any of its deps so it was relatively easy20:40
ToyKeeperrobru: Early warning, it looks like power use went up in the last couple images.  The increase coincides with the image when network-manager 1.2 landed, but I'm not sure if that's the cause.20:41
tedgOh, really, I was figuring that was on the bad list.20:41
tedgkenvandine: Can you see if copying pyflakes works?20:42
tedgIf not, I'll cut-and-paste the code I need.20:42
robruWell i dunno if we want pyflakes in the overlay ppa per se...20:43
tedgrobru: It's only a build-dep20:43
robruToyKeeper: can you file a bug? Thx20:43
ToyKeeperrobru: Yes, already doing so.  Just wanted to mention that things might get bumpy.20:43
robruToyKeeper: OK thanks20:43
* tedg puts his tray table in the stowed position20:44
robrutedg: can you make it conditionally a build dep? Like only used in xenial or later?20:44
kenvandinetedg, i can copy it... if we agree we should :)20:44
tedgrobru: Yes, we could.20:45
tedgThat's fine, I don't know that this will really be used much of vivid. More of a desktop feature anyway.20:46
kenvandinetedg, so you don't need it?20:46
robrukenvandine: do you have an opinion on pyflakes being in overlay? I guess it wouldn't ship on the phone, not sure about other implications20:46
tedgWait, how do I make a build-dep version dependent?20:46
kenvandinerobru, no opinion20:46
kenvandinetedg, ?????20:46
kenvandinei've seen you do that many times20:47
robrutedg: you'd have to use the pre processing si that it appears in xenial but not vivid20:47
kenvandineoh20:47
kenvandineyou mean series dependent20:47
tedgYeah, that's ugly.20:47
tedgWish there was a build-suggests20:47
kenvandineif we don't ship it, i don't mind copying it20:47
tedgkenvandine: We wouldn't, it is just used to check during the build of whoopsie20:48
kenvandinetedg, copied to silo 5720:48
kenvandinetedg, add it to the silo config20:48
tedgkenvandine: Will do, thanks!20:49
kenvandinenp20:49
kenvandinei need to go do kid things... so harass me on telegram if needed :)20:49
tedgWill do, if there's a longer chain here, I'm giving up on this line.20:49
* tedg is ready for the phone to be on Xenial already20:50
Saviqplars, no dice, it failed as soon as it started talking to it over ssh https://unity8-jenkins.ubuntu.com/job/test-ppa-autopkgtest/label=phone-armhf,package=unity8,release=vivid+overlay,testname=autopilot.sh/159/console - /me will add some debugging info20:51
robrutedg: yeah me too20:51
plarsSaviq: :(20:51
plarsSaviq: rick will be back tomorrow, maybe he can see if there's some physical issue going on with those devices20:51
Saviqplars, ack, I'll try and get a bit more info about it20:52
Saviqplars, what else do you think could be useful http://pastebin.ubuntu.com/16324468/ ?20:55
Saviqadding "adb forward --list"20:56
oSoMoNubuntu-qa: silo 38 has a fix for https://launchpad.net/bugs/1576639 , I know it’s past theoretical final freeze but I think the ratio benefit/risk is high, I’ve spent the day testing it extensively on 3 different devices, and if possible would like to have it included in OTA-11. Any chance of that happening?20:57
ubot5Ubuntu bug 1576639 in webbrowser-app (Ubuntu) "memory threshold is too high" [High,In progress]20:57
oSoMoNautopkgtests are currently running on that silo, it should be ready for QA validation soon20:58
Saviqplars, oh well, looking at syslog alone it's not looking great https://unity8-jenkins.ubuntu.com/job/test-ppa-autopkgtest/label=phone-armhf,package=unity8,release=vivid+overlay,testname=autopilot.sh/161/console20:59
Saviqsry, wrong link https://unity8-jenkins.ubuntu.com/job/test-ppa-autopkgtest/label=phone-armhf,package=unity8,release=vivid+overlay,testname=autopilot.sh/162/console21:00
robruoSoMoN: beg sil2100 i guess21:08
* oSoMoN bows and begs sil210021:09
oSoMoNI think pmcgowan and a bunch of others were keen on seeing that fixed in OTA11, too21:10
ToyKeeperoSoMoN: I'm probably not the right person to ask.  I don't even have a relevant device to try it on.21:11
oSoMoNToyKeeper, any touch device will do (and even a desktop setup with not too much RAM in it, for that matter, as it’s not touch-specific code)21:12
ToyKeeperoSoMoN: Regardless, you'll need to ask someone who can approve an exception.21:14
oSoMoNToyKeeper, understood, thanks21:14
* oSoMoN heads to bed for now21:15
plarsSaviq: it's worth checking, but honestly we don't normally see too much of use even in syslog21:39
Saviqplars, sure, just saying there's a lot of usb-related problems reported there21:40
Saviqplars, "Slave went offline" happens too often, too - and that's about heymann IIUC, not about the phone?21:41
Saviqhttps://unity8-jenkins.ubuntu.com/job/test-ppa-autopkgtest/label=phone-armhf,package=unity8,release=vivid+overlay,testname=autopilot.sh/162/console21:41
Saviqhappened twice today21:41
renaturvr, ping21:42
plarsSaviq: I do see that the jenkins slave restarted around that time, but there's no error or crash or anything, so it's not clear why21:47
Saviqplars, ack21:47
plarshmm, there's a phone I don't know about plugged into that system21:49
cjwatsontedg: you can always build-depends: thing-new-in-xenial | base-files23:59
cjwatsontedg: trade-off is that if it's transiently uninstallable or something then you may (or may not) find the build-dependency just gets (effectively) ignored when you wanted it to fail23:59

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