/srv/irclogs.ubuntu.com/2015/06/04/#ubuntu-ci-eng.txt

=== _salem is now known as salem_
=== chihchun_afk is now known as chihchun
plarscyphermox: I have found at least part of the problem, and I got a run of amd64 to pass. It doesn't seem to work everytime so it's still a little flaky but it's better. I'll try to do some more debugging on it tomorrow or pass it off to the next vanguard04:39
cyphermoxplars: ack04:40
cyphermoxwhat seems to be the problem?04:40
plarscyphermox: well, at first I was thinking it was the command used to restart rsyslog - it was using 'restart' rather than 'systemctl restart' as is likely needed by systemd now. But that didn't seem to fix it. I have it trying both now though, and not failing for that alone since we need to support older isos also04:45
plarscyphermox: part of the problem seems to still be permissions around /dev/ttyS0 though. That was changed a while back and it broke us. So we're adding syslog to the dialout group, setting permissions, etc. It still seems to persistently want to fail in some cases though. I'm not convinced there isn't some other error lurking there either, but I seem to be04:46
plarsable to work around it for now04:46
plarscyphermox: to be fair, I'm not really convinced of the usefulness of these tests, they are just doing installs with different combinations, and preseed ones at that. There are autopilot tests out there for driving ubiquity through most of these same install scenarios.04:47
plarsthat seems like a much better test for the desktop install process to me04:48
=== salem_ is now known as _salem
infinityplars: If you want to support both upstart and systemd, what you're looking for is "service $service restart"06:05
abeatoMirv, any idea why is that ^^ (nuntium not found in archive)07:48
abeatoit is a wily silo, that I was not able to publish before due to a bug in the gccgo powerPC compiler (now apparently fixed)07:49
Mirvabeato: err08:24
Mirvabeato: it's a sync silo trying to sync from silo 27, but that is already gone08:26
Mirvabeato: so it should probably be a sync silo from overlay PPA instead08:26
abeatoMirv, ah, I see... but there is a package in the PPA08:27
Mirvabeato: yes, but you're trying to rebuild which would mean trying to sync again08:27
abeatoMirv, anyway, which is the syntax I should use then?08:27
Mirvabeato: well I just put it in, reconfiguring now08:27
abeatoah, just seen, thanks08:27
Laneyfginther: the things you don't know ...08:28
Laneyfginther: Looks like Source can be Source: <source package> (version)08:28
Mirvthe spreadsheet is giving a bit of trouble though again08:28
Mirvok, now it orks08:28
Laneyfginther: so do grep-aptavail -e ... "^${SRC}( \(.*\))?$" or something08:29
Laneyfginther: and get rid of -X08:29
LaneyI think that should do it...08:29
Mirvabeato: reconfigured + started https://ci-train.ubuntu.com/job/ubuntu-landing-011-1-build/166/console08:29
abeatoMirv, awesome, thx08:30
Mirvogra_: I'm not sure if it went unnoticed, but see http://people.canonical.com/~ogra/touch-image-stats/20150521.changes - GTK2 dropped! :)10:03
seb128Mirv, do you know what changed to drop it? qt?10:04
Mirvseb128: "me" :)10:04
Mirvseb128: I mean, the gtk2 theme support was split into separate package10:04
Mirvafter Debian settled on the way to do it10:05
seb128Mirv, oh, great :-)10:05
Mirvthat was the easier thing (than dropping gtk3) but also needed to be done10:05
ogra_Mirv, heh, it actually went unnoticed because when i looked at it i was focused on libgo (which droped about 30MB from the image too)10:06
seb128right10:06
=== jamesh__ is now known as jamesh
Laneyhow do I delete a package from a landing?10:37
Laneyedit spreadsheet, delete from ppa, watch only build?10:38
Laneyah "reconfigure"10:43
jibeltedg, Hey, how can we test the fix in silo 30?11:21
jibelsince it happens on upgrade11:21
=== chihchun is now known as chihchun_afk
popeyfginther: are you mid way through fiddling with jenkins? Is that why this is failing https://code.launchpad.net/~carlos-mazieri/ubuntu-filemanager-app/samba-browsing-11/+merge/252982 ?12:00
=== alan_g is now known as alan_g|lunch
fgintherpopey, that branch is failing due to bzr conflicts12:03
popeyfginther: hmm12:05
=== tvoss is now known as tvoss|afk
abeatotrainguards, I have changed the target distribution for line 45 (silo 19), how can I reconfigure the silo?12:36
Mirvabeato: it needs to be freed up and reassigned. if that's ok (you lose the current builds), I can do that.12:40
abeatoMirv, that's fine, please do12:41
Mirvpete-woods: needs top approval https://code.launchpad.net/~pete-woods/indicator-network/15.10-fix-unity8-cycle/+merge/26084912:47
Mirvabeato: ^ 011 the dual silo12:48
abeatoMirv, awesome, thx12:48
pete-woodsMirv: fixed them both12:49
pete-woodssorry about that12:49
Mirvpete-woods: np12:49
jibelcharles, is silo 46 ready?13:07
tedgjibel, Honestly, I'm not sure.13:10
tedgjibel, In theory just upgrading the package and rebooting should be the same.13:10
tedgjibel, Then installing a new version of an app13:10
=== alan_g|lunch is now known as alan_g
jibeltedg, installing the silo on an upgraded device that shows duplicate icons would work?13:18
tedgjibel, No, you'd have to install or remove an application as well in that case.13:19
jibelok13:20
tedgjibel, My concern with this bug is that it was vivid for two months and no one saw anything. I think we've fixed a bug, but I'm not sure it's *the* bug. I'm not sure there's much we can do with the silo, but I do think that on the next image we need to really flag this to watch out for.13:21
pmcgowanjibel, silo 46 wfm, mind giving it a go13:40
pmcgowantedg, any leads on that dupe/missing icon thing13:40
jibelpmcgowan, works for me too, I was waiting for charles to confirm that it's ready to land13:41
pmcgowangreat13:41
tedgpmcgowan, silo 3013:41
pmcgowannice13:41
pmcgowantedg, how to test it?13:41
jibelpmcgowan, om26er_ is verifying silo 30 but it is not clear how to test it without building an image13:42
tedgpmcgowan, You basically need to install the silo and upgrade/remove an app. But… yeah, what jibel said.13:42
pmcgowanok will wait for omer then13:42
tedgpmcgowan, My big concern is that no one saw this in vivid for two months. So it makes me a bit worried about it.13:43
tedgI don't think everyone running vivid is wiping or something like that.13:43
pmcgowanyeah we install and update13:44
pmcgowantedg, I only saw it by reinstalling rtm then vivid afterword13:44
jibelpmcgowan, bug 1461952 looks serious too13:44
ubot5bug 1461952 in telepathy-ofono "SMS to multiple reciepients failed" [Undecided,New] https://launchpad.net/bugs/146195213:44
pmcgowanjibel, yeah let me get them looking at that13:46
pmcgowanbfiller, ^^13:46
jibeltedg, what is the risk of regression with silo 30?13:48
tedgjibel, Software always gets better! ;-)  I think it's pretty low. We're only changing the filters to look at more files. Worst case we delete a user defined desktop file, which is not a "normal user" by any means.13:50
pmcgowanjibel, for me it doesnt send the group message at all, just says failed13:52
jibeltedg, so an option would be to confirm that the silo works on OTA4 and doesn't regress anything even if we cannot confirm that it fixes the original upgrade issue, land it, build an image and redo the upgrade test.13:53
tedgjibel, +113:53
pmcgowaniahmad, for that sms bug, does it send to one of the recipients or just Failed?13:54
jibelfor me it doesn't send the SMS either, just 'failed' (untranslated btw) after a several seconds.14:00
charlesjibel, could I get a second pair of eyes on silo 46? It looks strange to me:14:01
charlesjibel, from the dashboard, the MPs look correct. The issue yesterday was a new MP was needed to only pull in the one new icon instead of tiheum's whole suru v2 branch14:01
charlesjibel, but if you look at the diff file in 46's PPA it looks like that build still pulled in all of suru v214:03
charlesmy guess is that this is a CI issue and wouldn't affect the actual landing, but would like a second opinion...14:04
iahmadpmcgowan, just failed14:07
pmcgowanack14:07
jibelMirv, ^^ any idea about charles question, why the diff in the PPA is not the same than the diff in the MP and in jenkins?14:08
dobeysounds like that silo needs reconfigured14:08
dobeyand then rebuilt14:09
om26er_tedg, so I downgraded to last promoted image, upgraded through system-image-cli to proposed and then install the silo packages and rebooted.14:11
om26er_I no longer see duplicate packages. tedg is that expected ?14:11
tedgom26er_, It's possible. It would happen anytime the click hook is run. Installing a package is how you could guarantee that, but there are other reasons they'd run.14:12
tedgom26er_, So that's why I suggest installing an app, just to be sure they were run.14:12
om26er_tedg, I was not able to find the changelog. How "dangerous" is this silo ?14:13
tedgom26er_, pansy. Sure one could kill you but it'd have to be frozen and thrown at high speed.14:14
om26er_"this" :)14:14
Mirvjibel: diff in the PPA may be diff to the previous version that happened to be in the PPA, not the actual diff. that's why we have the jenkins generating the diff in the first place, since PPA/LP can't be trusted on providing a correct diff.14:15
Mirvcharles: jibel: not opening that 13MB diff but maybe the diff is from the previous build that included the wrong things, so the diff shows how it's reverted again? the jenkins diff shows the actual diff from archive version.14:17
Mirvso the jenkins generated https://ci-train.ubuntu.com/job/ubuntu-landing-046-1-build/5/artifact/ubuntu-themes_content.diff is the real diff14:18
jibelcharles, the actual debdiff with the version in the PPA and the version on the phone is http://paste.ubuntu.com/11566558/14:21
jibelcharles, there is a change for gtk-3.0 and the rest if the green icon14:23
jibelno sure why it is slightly different from jenkins though14:25
balloonsfginther, confirming for myself and perhaps popey . We'll be shutting off the AP tests for now on core apps, yes?14:30
fgintherballoons, yes14:31
balloonsty.14:31
renatuballoons, I got approval from rsalveti on silo 16. Can we land it?14:34
charlesjibel, excellent, thanks for looking at that14:36
charlesMirv, thanks :-)14:36
charlesjibel, pmcgowan, then IMO silo 46 is ready for QA signoff wrt card https://trello.com/c/JXxJRGIt/1753-ubuntu-landing-046-unity8-ubuntu-themes-renatu-charles14:37
pmcgowancharles, sorry it was so much trouble14:38
charlespmcgowan, sorry it didn't land yesterday as it should have; I should have started off with a cherrypicked suru MP14:39
rvrawe: I attached the syslog14:44
om26er_tedg, so what else should I be testing? Can you help ?14:44
awervr, thanks much!14:45
tedgom26er_, There's not a lot more, really all it does is change the hook that creates the desktop files.14:45
tedgom26er_, If you look at ~/.local/share/applications things should look "right"14:45
tedgom26er_, i.e. no duplicate apps or apps missing.14:45
jibelom26er_, maybe additionally install and remove an app to check if icons are added/removed and there is not much else we can do14:51
jibelcharles, can you update column K on the spreadsheet for line 56?14:54
om26er_jibel, hmm, let me try that as well14:56
om26er_I was about to approve it14:56
balloonsrenatu, perhaps you meant boiko earlier?14:57
jibelom26er_, unless you find anything else useful to try from https://wiki.ubuntu.com/Process/Merges/TestPlan/ubuntu-app-launch it's fine to approve14:57
renatuballoons, boiko is off for the rest of the week14:57
fgintherLaney, thanks for the help. That should help get gtk-3.0 unwedged14:59
=== chihchun_afk is now known as chihchun
pmcgowanjibel, can you get the content of .cache/upstart/nuntium.log15:11
jibelpmcgowan, http://paste.ubuntu.com/11567396/15:13
pmcgowanjibel, yeah same as here15:14
pmcgowanwait yours is diff15:15
jibelpmcgowan, are group sms sent as mms?15:15
pmcgowanjibel, yes, seems no mms is working on krillin15:15
jibelpmcgowan, ah it might fail because wifi is on15:16
pmcgowanoh15:16
pmcgowanstill fails15:16
pmcgowankenvandine, see his paste15:17
jibelpmcgowan, I can send with wifi off15:17
jibelit's confusing15:18
pmcgowanreally15:18
kenvandineoh... there was a fix for sending with wifi on just recently15:18
kenvandineabeato, ^^15:18
jibelbug 141797615:18
ubot5bug 1417976 in nuntium (Ubuntu) "Cannot send MMS with Free Mobile when on Wifi" [High,Confirmed] https://launchpad.net/bugs/141797615:18
abeatojibel, which is the output of "ip route" ? which image are you using?15:19
jibelabeato, with wifi on or off?15:20
abeatoon15:20
charlesjibel, sorry was in meeting, yes will update column K on spreadsheet line 5615:20
jibelabeato, I'm using rc-proosed/bq-aquaris.en #24 aka OTA415:20
abeatojibel, also, which is your operator?15:21
jibelabeato, http://paste.ubuntu.com/11567572/15:22
jibelabeato, Free Mobile15:22
abeatojibel, cout you paste "cat /etc/NetworkManager/dispatcher.d/03mmsproxy" ?15:23
abeatojibel, and output of execution of /usr/share/ofono/scripts/list-contexts15:24
nerochiaroelopio: hello, i am having different results running a test in autopilot3-sandbox-run and normally. do you know of any known bugs in sandbox-run that might cause this ? the tests have something to do with keyboard shortcuts15:25
nerochiaroelopio: on webbrowser15:25
jibelabeato, http://paste.ubuntu.com/11567770/ + http://paste.ubuntu.com/11567824/15:28
abeatocyphermox, awe I remember to have read in some comment in some bug a problem when creating routes in case the proxy was empty in 03mmsproxy15:28
abeatomight be related to jibel 's issue?15:29
abeatojibel, thanks15:29
jibelyw15:29
aweabeato, I don't recall seeing that...15:30
abeatojibel, one more thing :)... what do you get when doing "grep mmsproxy /var/log/syslog" ?15:30
aweabeato, did you verify he has the updated mmsproxy dispatcher script?15:30
awebrb15:31
abeatoawe,yes that's fine, it is not a combined context, but proxy is empty for him15:31
awewhat do you mean by proxy is empty?  No MessageProxy defined in the context?15:32
abeatoawe, right, see http://paste.ubuntu.com/11567824/15:33
* awe looks15:33
aweabeato, is the messagecenter hostname resolvable & reachable over the WiFi network?15:34
jibelabeato, http://paste.ubuntu.com/11567902/15:34
abeatojibel, thanks, it is this issue with the proxy, free does not define it and causes trouble with the routes15:35
abeatoawe ^^15:35
aweindeed, I bet the mmsproxy script doesn't explicitly check for no proxy?15:37
abeatoyes15:37
aweanyways, we need to determine (a) is the hostname resolvable over the WiFi network15:37
aweand (b) is the resulting IP address reachable over the same15:37
aweit could be that we need addtional routing logic added if there's no MMS proxy defined15:38
abeatoprobably it is the case15:38
awejibel, can you see if you can resolve the messagecenter hostname when connected to WiFi15:40
awe?15:40
aweand then see if the resulting IP address is reachable?15:40
awealso, is there a bug for this?15:40
aweabeato, ^^15:40
abeatohttps://bugs.launchpad.net/ubuntu/+source/nuntium/+bug/141797615:40
ubot5Launchpad bug 1417976 in nuntium (Ubuntu) "Cannot send MMS with Free Mobile when on Wifi" [High,Confirmed]15:40
aweright15:40
awethanks15:41
aweabeato, can assign to you?15:41
abeatoindeed15:41
pmcgowanI think I disconnected, so15:42
pmcgowan<pmcgowan> abeato, awe mms is not working for me on AT&T on either arale or krillin15:42
pmcgowan<pmcgowan> https://bugs.launchpad.net/canonical-devices-system-image/+bug/146195215:42
ubot5Launchpad bug 1461952 in Canonical System Image "SMS to multiple recipients failed" [Critical,Confirmed]15:42
awepmcgowan, ;(15:42
awewith ota4?15:42
awedo you have an active data plan?  maybe they cut you off too!15:43
pmcgowanawe, heh15:46
pmcgowanyes ota415:46
pmcgowanawe, I get internet15:47
abeatopmcgowan, please paste ~/.cache/upstart/nuntium.log15:47
pmcgowanabeato, this is arale http://pastebin.ubuntu.com/11567998/15:48
pmcgowanabeato, this is krillin http://paste.ubuntu.com/11567348/15:48
pmcgowansame sim15:48
awepmcgowan, my data plan is empty; I'm going to check with AT&T, but I can't reproduce while just on mobile data15:49
awepmcgowan, in this case, when I try to send the MMS, the messaging-app just continues to show the spinner for the message15:49
aweI never see it stop and indicate failed15:50
charlesjibel, line 56 marked as tested15:50
awealso, network-indicator isn't showing me connected to WiFi when I am15:50
awe;(-15:50
awepmcgowan, do you see the same endless spinner in the messaging-app?15:50
awebfiller, do you remember seeing that at the sprint in austin?15:51
* awe my camera seem super-unstable on arale15:52
rvrCan someone confirm this with image 24? https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/146199915:56
ubot5Launchpad bug 1461999 in network-manager (Ubuntu) "Crash changing SIM slot technologies" [Undecided,New]15:56
jibeltedg, can you mark column k of line 57 as tested (if it's tested of course) so the silo can publish15:58
tedgalecu, ^ did you test that silo?16:00
abeatopmcgowan, you do not have cellular data activated for krillin16:00
abeatopmcgowan, for the moment we need tht16:00
pmcgowanabeato, I don't? I can try again in a min16:02
abeatopmcgowan, the contexts listed in the nuntium log are all inactive16:02
pmcgowanweird16:03
awepmcgowan, can you also dump the output of /usr/share/ofono/scripts/list-contexts?16:06
aweit appears that the apn db update introduced a new APN ( nextgenphone ) for AT&T16:07
aweyou should now see ATT Phone, ATT WAP, and ATT Nextgenphone16:07
awethe latter fails for me if WiFi is enabled16:07
pmcgowanabeato, ok basic mms does work there now, so I am confused16:08
pmcgowancell data was off as you said16:08
pmcgowanbut group mms still fails16:08
awepmcgowan, abeato and I discussed this yesterday; the messaging-app should display an error if mobile-data is off16:09
abeatopmcgowan, for arale, in your pastebin I see:16:09
abeato2015/06/04 15:26:02 m-send.conf ResponseStatus for 6f58b19377ad081d8c67309452ee53c9 is 22616:09
abeatoError-permanent-message-format-corrupt = <Octet 226>16:09
abeatopmcgowan, was that when sending a group message?16:09
aweat least for the short term.  abeato has a proposed auto-enable for MMS, but we need to focus on short-term message stability first and foremost16:09
aweabeato, does group SMS use MMS under the covers16:10
abeatoyes afaik16:10
aweI had no input in it's design or impl16:10
pmcgowanawe, yes group chat needs mms16:13
pmcgowanabeato, give me a min, on a call16:13
abeatook16:13
pmcgowanabeato, http://pastebin.ubuntu.com/11568609/16:14
pmcgowanthats a group send on krillin at&t16:14
abeatopmcgowan, taking into account that non-group MMS is working, and being error 226 a bad format error, we must be sending something in the payload that att does not like16:17
abeatopmcgowan, I think the data is prepared by messaging-app16:17
abeatopmcgowan, nuntium wraps it and sends16:17
=== chihchun is now known as chihchun_afk
jibeltedg, if alecu is not online can you do it, we really need this fix then rebuild an image16:20
om26erjibel, tedg approved16:23
jibelom26er, tedg ah thanks I missed it16:24
jibelom26er, no it is not16:24
jibelom26er, line 5716:25
om26erjibel, yes, says Granted16:25
om26erjibel, refresh your sheet :p16:25
jibelom26er, right for QA, but column K is not set to tested16:26
om26eroh16:26
awepmcgowan, I just restored my AT&T SIM, billing error on their end.  I'll retry MMS after lunch16:26
fgintherpopey, balloons, music-app and ubuntu-clock-app should now be un-blocked16:26
* popey hugs fginther 16:26
tedgjibel, I'd really prefer alecu to do it, I think he's probably back soon (and I'm about to head out). He had an upgraded OTA 3.5 device he was going to test it on.16:27
jibeltedg, ok16:27
tedgjibel, But if he's gone for longer I can test it after I get this other silo done.16:27
tedgNot in a clean state right now.16:28
alecutedg: I'm back from lunch, I'll give it a try now16:34
alecutedg: but it's silo 030 that I should test, right? not 008 as suggested above16:37
renatuballoons, are you a trainguards? could you release the silo 16?16:44
renatuI got the ubuntu core dev approval16:44
balloonsrenatu, I'm not a trainguard16:44
balloonstrainguards ^^16:45
renatuballoons, ok sorry :D16:45
balloonsno worries, sorry if I misled you :-)16:47
=== alan_g is now known as alan_g|EOD
Mirvrenatu: it needs an archive admin ack for the new binary package and I haven't gotten one on #ubuntu-release. robru can continue to monitor it (and robru MOTU ack on other packaging changes..)17:03
Mirv(= I ack the other changes)17:04
robruMirv: actually with the new dual landing support, it's now possible to change the target series without freeing and reassigning17:30
alecutedg: installed silo 030 on my ota-3.5. Then upgraded to ota-4 (with the steps outlined by om26er): I get duplicated icons17:34
alecutedg: I suspect this has to do with the ota-4 overwritting stuff from the silo17:35
=== tvoss|afk is now known as tvoss
om26eralecu, is there a new image ?17:35
alecutedg: any hints on how to test this properly?17:35
alecuom26er: I think there's not. So, the current ota-4 image steps on whatever I installed from the silo17:36
alecuI'll try installing the silo on top of ota-4, but I still have no clue on how to test it properly.17:39
jibelalecu, from earlier discussion, I think there is no good way to test it without an image and the best we can do is to ensure the silo doesn't regress anything.17:41
jibelalecu, 3.5 to 4 is a full image upgrade and the silo will be overwritten17:41
alecujibel: right. Then I'm not sure I'm the right person to test this silo; I offered ted help testing this issue, but I've never tested this whole package.17:43
alecujibel: I can give it a go if nobody else is around17:43
alecujibel, om26er: btw, good news are that installing the silo with the citrain script on top of ota-4 (and rebooting), shows only one copy of each of the icons17:45
om26eralecu, yeah, thats what I noticed as well17:46
pmcgowanjibel, so in the end I actually have group chat working on my krillin17:47
jibelpmcgowan, good, me too with wifi disabled. I suppose the original reporter's issue is also related to MMS configuration.17:50
pmcgowanjibel, could be, will downgrade severity17:50
jibelalecu, if you and om26er tested it and didn't find any obvious breakage, it's good to land. what do you think?17:51
jibeltrainguards, can you publish silo 46?17:52
jibelrobru, ^17:52
robrujibel: done. so that'll be in overlay immediately and wily-proposed shortly17:53
jibelrobru, thanks17:53
robrujibel: you're welcome17:54
alecujibel: sounds good to me. I'm finishing with the test plan for that silo, looking good so far17:55
jibelalecu, perfect, thanks.17:55
alecujibel: the tests from ubuntu-app-launch/security-app-launch are taking forever. The rest is +118:12
alecujibel: and just now they are finished. I'm +1 the silo in the spreadsheet18:13
alecujibel: om26er, tedg: ^18:15
alecuyay!18:37
tedgGreat, thanks alecu!18:38
alecutedg: thanks to you, sir18:39
robrulol, every time19:28
robrualexabreu: silo 2819:43
alexabreurobru, thx !19:44
robrualexabreu: you're welcome!19:44
=== tvoss is now known as tvoss|eod
=== tvoss|eod is now known as tvoss

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