/srv/irclogs.ubuntu.com/2014/12/01/#ubuntu-ci-eng.txt

imgbot=== trainguards: IMAGE 39 building (started: 20141201 02:05) ===02:05
=== chihchun_afk is now known as chihchun
=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
imgbot=== trainguards: IMAGE 39 DONE (finished: 20141201 03:25) ===03:24
imgbot=== changelog: http://people.canonical.com/~ogra/touch-image-stats/39.changes ===03:24
=== chihchun is now known as chihchun_afk
Mirvindeed06:24
=== chihchun_afk is now known as chihchun
=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
Mirvogra_: do you still do the occasional touch seed upload? nik90/bzoltan would like to get the one-liner https://code.launchpad.net/~nik90/ubuntu-seeds/add-qml-connectivity/+merge/237442 in although that branch happens to be against utopic.09:25
ogra_Mirv, yeah, i saw the comment on the branch09:25
alan_gcihelp is something broken? We've not seen a run complete since before the weekend. Here: http://s-jenkins.ubuntu-ci:8080/job/mir-mediumtests-runner-mako/09:26
Mirvogra_: ok.09:28
alf__cihelp: Hi! A lot of Mir jobs have been blocked waiting for makos since Friday. Any idea what is wrong?09:40
alf__cihelp: e.g., http://s-jenkins.ubuntu-ci:8080/job/mir-ci/09:40
alf__alan_g: ^^ FYI09:40
=== vrruiz_ is now known as rvr
evhi alf__ and alan_g. We're looking into it now.09:46
alf__ev: great, thanks09:46
Saviqtrainguards ↑ please :)10:04
MirvSaviq: okie10:05
popeyogra_: bug 139791810:08
ubot5bug 1397918 in unity-scopes-api (Ubuntu) "krillin #169 + cgroups fix, scopes hang / crash / unusable" [Undecided,New] https://launchpad.net/bugs/139791810:08
* ogra_ hugs popey 10:08
psivaaogra_: does 'Dec  1 10:19:47 heymann kernel: [5817795.697394] usb 1-1.5.2.1: usbfs: USBDEVFS_CONTROL failed cmd fastboot rqt 128 rq 6 len 254 ret -108' ring any bell?10:23
psivaathis is a kernel log from the adb host10:23
ogra_hmm, nope10:23
ogra_looks bad thouh10:23
ogra_+g10:23
psivaayea, i *think the adb host in the lab is somehow messed up.. failing to see how, yet10:24
ogra_hub issues ?10:24
michicihelp: cloud-worker-06 is running ridiculously slowly. It’s been like this since last Friday. I pinged about this back then, but didn’t get a response. The machine usually is still compiling by the time the arm build for the same code has compiled and finished testing.11:02
michicihelp: could someone please fix that builder or take it off-line?11:02
=== greyback_ is now known as greyback
Saviqcihelp, also, looks like makos are dead http://s-jenkins.ubuntu-ci:8080/job/generic-deb-autopilot-runner-vivid-mako/11:17
Saviqthey all got stuck flashing11:17
psivaaSaviq: yea, working on them11:17
Saviqpsivaa, ok11:17
alan_gev: is this hung on the same issue? http://s-jenkins.ubuntu-ci:8080/job/qtmir-ci/11:29
michicihelp: Our build on cloud-worker-06 has been running for nearly and hour and still hasn’t finished compiling. Can someone PLEASE look at this?12:02
michihttp://s-jenkins.ubuntu-ci:8080/job/unity-scopes-api-devel-vivid-amd64-autolanding/34/console12:03
psivaamichi: just finishing up some mako device issues in the lab. will take a look after that12:03
michipsivaa: Thank you, I appreciate it!12:04
psivaaSaviq: alf__: alan_g: mako devices have a broken image (34) and failing to comeback to fastboot mode. we need someone onsite to take a look and flash them with a newer image. i've reported a ticket for that12:14
Saviqpsivaa, ktx12:15
alan_gpsivaa: thanks12:15
psivaamichi: cloud-worker-06 has been taken offline now, the job is building now on -11, please let us know if you see that being slow too12:16
michipsivaa: Awesome, thank you!12:16
michiWhy do these things suddenly start running slowly?12:16
psivaamichi: haven't digged in to the instances yet, will do in a bit12:16
michiOK, cool. I’m not trying to be a pain. Just curious.12:17
psivaanp :)12:18
psivaaogra_: wifi in the lab also *appears broken, could not come to any other conclusion, a device with image 38, which was confirmed to be working also now failing to connect to the network. i've reported a ticket for that too with the IS12:19
Saviqtrainguards, Icanhasreconfigure of vivid silo 3, had to add settings because of a missing dependency there12:22
ogra_psivaa, ouch12:24
MirvSaviq: done12:33
SaviqMirv, thanks12:34
Ursinharsalveti: you might like this: http://lizkeogh.com/2008/09/10/feature-injection-and-handling-technical-stories/12:46
* rsalveti checking12:47
jgdxxcihelp: Hi there. Any idea what this test failure [1] really means? [1] https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-vivid-mako/338/testReport/junit/ubuntu_system_settings.tests.test_sound/SoundTestCase/test_sound_page/14:22
=== jgdxx is now known as jgdx
fgintherjgdx, The test runs, but it looks like it couldn't find some widget. I don't have much more advice on this, but if you want to try to reproduce exactly as jenkins runs the job, there are some starter instructions at the bottom of this file: http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/README-cli.rst14:35
jgdxfginther, thanks14:46
fgintherjgdx, just one note. running those instructions as-is will flash your device (that should be included in the README)14:47
=== chihchun is now known as chihchun_afk
=== oSoMoN_ is now known as oSoMoN
Mirvogra_: ^ I guess we're still under no-landings rtm mode until otherwise told?15:01
=== plars changed the topic of #ubuntu-ci-eng to: Need a silo? ping trainguards | Need help with something else? ping plars | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: RTM Archive frozen (no new silos landing) ! RTM cron builds disabled
ogra_Mirv, if QA signed off i would expect that they chacked there are only ota-1 tagged bugs fixed in the silo ... but yeah, we should wait til we had a bug meeting with mgmt15:03
Mirvwhen the gates finally do open, the interesting thing is going through all the non "ota1" but critical rtm bugs15:04
Mirvogra_: yep, they probably check that but the official word is indeed closed still15:04
Mirvlike the UITK fixes some critical touch-2014-11-xx bugs but is not ota-1. lots of triaging to do.15:04
ogra_i would like a separately built image with silo3 landed first ... imho then we can open the gates again15:05
Mirvmakes sense15:05
=== chihchun_afk is now known as chihchun
alex-abreuMirv, mmh could you merge L46 & 47 in silo 5 ? :) there is no clear dep between the 215:10
alex-abreuMirv, thx :)15:12
Mirvalex-abreu: yw :)15:12
* Mirv out15:17
=== karni is now known as karni-afk
=== alan_g is now known as alan_g|tea
jhodappMirv, how are we handling getting a change that has already landed in vivid into rtm? We need a new rtm silo I'm sure, but then what do I need to do to get the particular revision of my branches into that silo?15:46
ogra_jhodapp, you need to make sure that your landing *only* closes the rtm ota-1 tagged bug15:46
ogra_i.e. no bundled fixes15:46
jhodappogra_, right15:48
jhodappogra_, but what needs to happen to do that? request a sync, propose a new MR, what?15:48
ogra_depends on the code ... if you have an rtm MR ... just request a landing15:49
jhodappogra_, I don't have an rtm MR15:49
ogra_if a sync is possible (because your fix went in standalone and without any other fixes in vivid) you can surely ask for a sync15:49
jhodappogra_, it did, there's a specific revision number for qtubuntu-camera that only contains my fix15:50
ogra_jhodapp, right, then a sync should be possible15:50
jhodappogra_, the trickier part is the android/hybris side, but those should have gone in cleanly too (I hope)15:51
ogra_yeah, that has to go through the normal john-mcaleely approval process15:51
jhodappogra_, right15:51
jhodappogra_, so how would that sync work? there was one change that landed in vivid for qtubuntu-camera after my fix...so how do we sync just my fix and not the revision change after mine?15:54
=== karni-afk is now known as karni
ogra_jhodapp, that would be a sil2100 question ... but he is sick today15:59
jhodappogra_, ok, I bet robru could answer too, no?15:59
ogra_i assume citrain allows syncing a specific version into a silo though15:59
brendandogra_, what exactly is wrong with the current vivid image?16:08
ogra_brendand, nothing that i know about ... do you have an issues ?16:08
ogra_*any16:08
brendandogra_, phablet-click-test-setup is getting stuck16:09
ogra_ah, havent heard of that one yet16:09
brendandogra_, did the tests run in ci today?16:10
ogra_brendand, nope, there was an issue with the wlan AP in the lab16:10
ogra_they couldnt finish provisioning due to that16:11
brendandogra_, ok16:11
brendandogra_, so phablet-click-test-setup hasn't been tested on 39 then16:11
ogra_do you know where phablet-click-test-setup hangs actually ?16:11
brendand'Checking out lp:music-app to /tmp/tmpd0Ftqv/work'16:11
brendandogra_, not sure if that's on the device or locally16:11
ogra_sounds like device16:12
brendandogra_, yeah seems to be the device16:12
ogra_put a "set -x" at the top of the script16:12
ogra_that should make it a bit more noisy16:12
brendandogra_, ah it's python16:13
ogra_ah, well ...16:13
=== alan_g|tea is now known as alan_g
brendandogra_, looks like there was a temporary launchpad issue16:19
oSoMoNtrainguards: hey, can I have silos for lines 48 and 49, pretty please?16:21
ogra_brendand, phew16:25
sergiusenscan people leave the sheet for a bit? "Wow, this file is really popular! It might be unavailable until the crowd clears. Try again."16:28
sergiusensah, there we go16:29
ogra_it's popular :)16:29
sergiusensmeh, now it can't save... :/16:31
ogra_robru, are u around ?17:06
=== plars changed the topic of #ubuntu-ci-eng to: Need a silo? ping trainguards | Need help with something else? ping cihelp | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: RTM Archive frozen (no new silos landing) ! RTM cron builds disabled
robrujhodapp: ogra_: wut? CI Train's sync code just grabs whatever package is in a distro and rebuilds it for the other distro. If you want to cherry pick only a certain fix you have to grab the specific diff and apply it as a new commit in an MP to thenew distro.17:28
ogra_robru, oh17:29
ogra_robru, i thought you could do something like copy-package of a sepcific version from the devel distro17:29
robruogra_: well you can totally do that but it has nothing to do with citrain.17:29
ogra_k17:29
robruogra_: so yeah, I can do a manual source upload silo and you can copy-package some older version into it I guess.17:30
ogra_right, i guess thats what jhodapp needs in his case ...17:31
=== greyback is now known as greyback|afk
=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
jhodapprobru, ogra_: yeah that's what I need17:56
robrujhodapp: do you have a spreadsheet row?18:03
jhodapprobru, yeah, 4718:03
robrujhodapp: what source package do you want synced? camera-app?18:04
jhodapprobru, it'd be qtubuntu-camera, hybris and the android package18:04
robrujhodapp: and what versions?18:05
jhodapprobru, for qtubuntu-camera: http://bazaar.launchpad.net/~phablet-team/qtubuntu-camera/trunk/revision/12018:06
jhodapprobru, not sure how the android and hybris changes would work18:07
robrujhodapp: well you'd have to go through the changelog and find the oldest version that has the fix you need18:08
jhodapprobru, yeah, will have to talk to rsalveti about this one since he always orchestrates these packages for landing18:08
* rsalveti looks18:09
ogra_will likely be hard to find an android package that doesnt ship any additional fixes18:09
jhodappyeah indeed18:09
jhodappthey're usually paired with several other things18:10
ogra_right18:10
rsalvetiyup, but should be fine, rtm has a quite recent android version18:10
rsalvetijhodapp: I can handle this landing18:10
ogra_we need to find a way to separate them out for future laandings18:10
rsalvetithere's no easy way18:10
jhodapprsalveti, ok, I'll be curious to know what you do though just so I can get an idea18:10
rsalvetibut that's for like every project18:10
ogra_well, cant we shelve them in git or some such18:11
rsalvetijhodapp: in this case we just need to sync qtubuntu-camera, libhybris and android from vivid into rtm, but we need to rebuild at least libhybris and qtubuntu-camera when doing so18:11
rsalvetican't necessarily rebuild android that easily18:11
ogra_so that you can just pull this one commit into an rtm package18:11
robrursalveti: ok well you have silo rtm 12, feel free to push whatever packages there18:11
rsalvetithen we need to push the hybris patch at the krillin git repo for rtm, and sync that landing with a device tarball18:11
jhodapprsalveti, yeah pretty involved18:12
rsalvetiogra_: we can, but in this case it's a pita to coordinate18:12
jhodapprsalveti, there's a Trello task for this, you can add yourself to it18:12
ogra_well, it wont pass QA  or management if there is anything else in the package18:12
rsalvetijhodapp: cool, will do18:12
rsalvetiogra_: well, that's something we need to check18:12
ogra_right18:12
rsalvetionce we understand if that will indeed be the case18:12
ogra_and to plan for the future18:13
jhodappogra_, this is really about change management and is always a hard problem at every company I've worked at18:13
rsalvetiwork to cherry-pick sometimes might be more than actually testing an additional fix18:13
ogra_jhodapp, yeah18:13
rsalvetithere's no silver bullet18:13
rsalvetionly manual work18:13
jhodapprsalveti, agreed...for qtubuntu-camera there's only one other fix on top of mine18:13
ogra_rsalveti, well, in case of android you only brick mako :)18:13
ogra_the landing for krillin would be a cherry pick in git anyway18:14
rsalvetijhodapp: right, that might be easy to cherry-pick18:14
jhodappshould be18:14
rsalvetibut then I need to manually sync that18:14
rsalvetiotherwise the train will bump the changelog18:14
ogra_so for the android package we could perhaps go as is18:14
rsalvetiand confuse everyone18:14
rsalvetithis is going to be painful18:14
ogra_but i think we need a general plan for this kind of stuff18:15
ogra_yeah18:15
rsalveti(not only this landing, as I know what I'm doing, but coordinating that with everyone)18:15
jhodappogra_, agreed18:15
rsalvetican't make train to rebuild stuff, etc etc18:15
ogra_rsalveti, yup ... and it wont get easier :/18:15
ogra_why cant you ?18:15
rsalvetiand only a few people has access to dput18:15
rsalvetiogra_: otherwise the version will be mixed18:15
rsalvetiI can have 2 bzr branches, vivid and rtm18:15
ogra_the train gets an older upstream package than vivid has already ... so it shouldnt break if it attaches rtm1 or whatever it does18:16
rsalvetiand have mrs landing on each one, separately18:16
rsalvetiogra_: right, but if I rebuild that, it'll bump the changelog18:16
rsalvetiwith today's date18:16
rsalvetiwhich will confuse everyone :-)18:16
ogra_for the upstream ?18:16
rsalvetithen it'll make it really hard to coordinate cherry-picks18:16
rsalvetiogra_: no, for the rtm branch18:16
rsalvetiso we can have qtubuntu-camera-(last week) on vivid and qtubuntu-camera-(today) on rtm, after a rebuild18:17
rsalvetibut the content will be different18:17
rsalvetiso the upstream will need to manually coordinate the cherry-picks18:17
rsalvetior the developer will need to take care of the changelog18:18
ogra_yeah18:18
* ogra_ prefers the changelog approach 18:18
ogra_technically we should take rtm packages and just add stuff on top of them18:18
=== alan_g is now known as alan_g|EOD
rsalvetiright, but that is only clear for us18:18
rsalvetiit can get messy really fast18:18
ogra_right, we should have some planning meetings with all parties involved18:19
rsalvetiyup18:20
ogra_on how to handle your stuff in vivid bset18:20
ogra_*best18:20
ogra_so that you dont step on your own toes18:20
robruis anybody going to land anything in vivid today? I have a from-scratch rewrite of the citrain publisher job I'd like to test out...18:20
=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
robrusergiusens: vivid 9 and 1019:57
sergiusensthanks19:58
=== barry` is now known as barry_
=== slangase` is now known as slangasek
=== barry_ is now known as barry
=== michihenning is now known as michi
=== Trevinho_ is now known as Trevinho
=== robru_ is now known as robru
=== psivaa_ is now known as psivaa
robrusergiusens: Saviq: rsalveti: anybody expecting to land silos today?21:57
Saviqrobru, no, tomorrow21:58
robruSaviq: ok, no worries.21:58
rsalvetirobru: sergiusens might21:59
robrursalveti: cool. he can be the guinea pig for the new publish job then ;-)21:59
sergiusensrobru: I will, but need to settle first23:20
robrusergiusens: no rush23:22

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