robrurenatu: ok we're back online if you need to build anything now00:45
=== robru changed the topic of #ubuntu-ci-eng to: Need a silo or CI Train support? ping trainguards | Need help with something else? ping cihelp | Train Dashboard: http://bit.ly/1I2YFOy | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: ci-train.ubuntu.com will be taken offline at 2200UTC for maintenance
renaturobru, thanks00:47
robrurenatu: you're welcome00:48
=== robru changed the topic of #ubuntu-ci-eng to: Need a silo or CI Train support? ping trainguards | Need help with something else? ping cihelp | Train Dashboard: http://bit.ly/1I2YFOy | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: Spreadsheet Replacement Rollout largely successful just dealing with some firewall hiccups.
robruAlbertA: hi, thanks for being among the first to try out the new train ;-) looks like there's a couple firewall issues stopping things from working, I've notified IS, hopefully will be resolved within an hour01:19
AlbertArobru: no prob. new site looks great!01:21
robruAlbertA: thanks!01:21
bregmarobru, is there any change to landing a silo once the testing has passed?01:43
robrubregma: nope that part is the same for now. you need QA to approve (sometimes) and then you need me to hit publish01:44
bregmarobru, ok, just making sure01:44
robrubregma: yeah the goal is to make it so that people can publish their own stuff too, but that's a ways off still01:45
robruAlbertA: ok wow, apologies on that horrible delay! retrying your build now, should be working: https://ci-train.ubuntu.com/job/ubuntu-landing-009-1-build/329/console05:16
=== chihchun_afk is now known as chihchun
robrumichi: sorry about that dput failure, retried for you, should be working now: https://ci-train.ubuntu.com/job/ubuntu-landing-051-1-build/14/console06:32
michirobru: Hey, no problem! Thanks for picking it up on it. I figured that, with everything in flux, I’d just wait a while for things to settle down :)06:33
michiThere was little point in shouting at you guys, I thought :)06:33
robrumichi: yeah that was a scary 8 hours of downtime that should only have been 2 hours of downtime. but as far as I know everything is perfect now, so have at it ;-)06:33
michiThe new stuff looks nice.06:34
michiI haven’t gone through the process of starting a new landing yet.06:34
robrumichi: thanks! please email me with any improvement ideas you may have. I'm going to try to focus on papercuts over the next week or so06:34
michiBut it sure looks easier than the spreadsheet.06:34
michiWill do, if I come across something that isn’t obvious.06:34
robrumichi: alright, 11:30 here, goodnight ;-006:34
michiWell-deserved sleep coming up. Enjoy it! :)06:35
jibelrobru, hey, just to confirm, the data from the spreadsheet won't be migrated to bileto?07:30
robrujibel: correct07:30
robrujibel: well maybe some of it, by hand, but generally no, they will coexist for a while07:31
jibelrobru, that's fine, it's just to know where we should approve landing requests for cards already created.07:32
robrujibel: good luck! Email me if you have any questions about the api07:33
jibelrobru, it should be all right, the API is easy to use and clear. Much simpler than scanning a spreadsheet :)07:34
robrujibel: hehe yeah.07:35
robrujibel: OK I'm just winding down, goodnight!07:35
jibelgood night07:36
jameshsil2100: fyi, we came up with a workaround for the thumbnailer gstreamer issues, so should be able to do a trunk landing with the gcc 5 fixes soon.07:58
sil2100jamesh: excellent, would like that ASAP07:59
jameshcihelp: I got this error from the generic-land portion of an autolanding job: http://s-jenkins.ubuntu-ci:8080/job/generic-land/27514/console08:00
psivaajamesh: let me take a look08:00
seb128jamesh, hey, is there a bug about that gstreamer issue?08:01
jameshpsivaa: everything built fine on the arch specific subjobs that preceded that bit08:01
jameshseb128: I haven't yet.  I believe the bug is tied to the fdsrc element though, which has given us problems in the past08:02
jameshseb128: I changed the thumbnailer to not depend on that element any more.08:02
seb128jamesh, would be nice to have a bug, forwarded upstream if possible08:02
jameshseb128: right.  I've done that for a number of fdsrc bugs in the past, but hadn't narrowed this one down.08:03
jameshpsivaa: if it helps, the MP for that landing is merging a copy of lp:thumbnailer to lp:thumbnailer/devel.  I wonder if this is some edge case for branch stacking, since there are no revisions in lp:/~jamesh/thumbnailer/devel-sync-20150725 that are not in the development focus branch08:06
psivaajamesh: that could be a possibility, i have seen this error once and trying to remember/dig for that information08:08
oSoMoNjibel, hey, I marked silo 6 ready for QA validation a while ago, and it hasn’t appeared in the trello board yet, known issue?08:12
jibeloSoMoN, I disabled the bot last night while robru was migrating the spreadsheet, just in case.08:13
jibelLet me run it08:13
jibelah, and robru moved the dashboard so now it crashes because it cannot find it08:15
jibelI'll fix that08:15
robrujibel: new dashboard is a bit different08:16
jibelrobru, where is the raw data?08:17
davmor2robru: grrrrrrr changing all the urls ;)08:17
robrujibel: requests.ci-train.ubuntu.com/static/json/index.txt lists assigned silos08:17
jibeloSoMoN, it'll take a moment08:19
jibellooks like we'll create cards manually this morning08:19
oSoMoNjibel, that’s alright, that one’s not super urgent08:19
psivaajamesh: would there be any possibility that there was a rev manually pushed to the devel branch08:19
jameshpsivaa: no.08:20
robrujibel: the filenames listed in index.txt can then be accessed relative to that url for the raw json statuses08:20
jibeloSoMoN, thanks, it'll be fixed soon. I didn't plan to make the bot work with the spreadsheet and bileto at the same time :)08:20
jibeloSoMoN, BTW, on the dashboard the state is 'Package Built' not 'Ready for QA'08:22
jibelit'll be a quiet day, nothing is ready for QA08:23
jibelrobru, ^08:23
oSoMoNjibel, then there’s an issue with the dashboard. robru: I marked silo 6 ready for QA validation in the spreadsheet earlier this morning, and apparently the dashboard didn’t pick it up08:23
sil2100The dashboard doesn't seem to fetch statuses from the spreadsheet at all now08:24
jameshpsivaa: the revision ci-train-bot@canonical.com-20150715020328-bq2ohz4ufnerale2 it says is a ghost is the head revision of lp:thumbnailer (and also of the devel-sync branch that is the source of the merge)08:24
robruoSoMoN: the dashboard no longer cares about the spreadsheet08:24
jibelrobru, so how do we track statuses during the transition?08:24
robrujibel: you'd have to look at the spreadsheet directly08:24
jibelrobru, of course08:25
robrujibel: sil2100: yes the transition may be a little bit rougher than i anticipated, to many moving pieces08:25
robrujibel: the trick is that the spreadsheet won't automatically update statuses either, so manual poking will be necessary to make it actually say "ready for qa"08:26
pete-woodsso the spreadsheet lives on?08:27
jibelit's like picking up your phone every 10 minutes just in case someone calls you because the ring is broken08:27
pete-woodsbut behind the scenes now?08:27
jibelno way we scan the spreadsheet manually08:27
psivaajamesh: i see a similar issue in this lp bug, https://bugs.launchpad.net/bzr/+bug/1161018/comments/408:33
ubot5Ubuntu bug 1161018 in Bazaar "bzr revno fails on ghost ancestry" [High,Confirmed]08:33
psivaawould you mind trying the workaround in comment #408:33
pete-woodsah right, okay just transitional :)08:34
robrujibel: we may need to tell people to just irc ping each other until the spreadsheet fully dies08:34
jameshpsivaa: sure.08:35
jameshpsivaa: If I have any trouble after retrying the landing, I'll ask again.  Thanks.08:37
psivaajamesh: sure, i tried it after copying your branch and it works.08:37
psivaajamesh: to test,  'bzr revno lp:~jamesh/thumbnailer/devel-sync-20150725' command should just work08:38
jameshpsivaa: pushed an empty commit, and "bzr revno" works, so hopefully Jenkins will be happy.08:40
psivaajamesh: http://s-jenkins.ubuntu-ci:8080/job/generic-land/27517/console was happy :)08:43
jibelrobru, yeah, there is no other choice anyway08:50
jameshpsivaa: thanks for your help.08:50
psivaajamesh: np08:50
jibelsil2100, robru don't you think that silos on the spreadsheet that must be rebuilt or reconfigured should be moved to bileto instead?08:54
sil2100jibel: that's one proposition, yes08:54
jibelsil2100, and every team is responsible of moving their records08:54
sil2100We'll probably doing that, but currently it requires some manual work...08:54
jibelsil2100, otherwise the spreadshhet will stay there forever08:55
sil2100Manual work not really suited for normal team members08:55
jibelsil2100, checking lines manually also requires manual work08:55
sil2100I know, just saying it's not something normal landers can do, we'll be migrating entries slowly ourselves08:57
sil2100I didn't expect we'd have to be doing that as a priority, thought that the dashboard would support both during the grace period09:02
jibelsil2100, if a line is copied from the spreadsheet does it mean the silo will have to go through the whole process again, merge, build, ...09:05
sil2100No, it won't have to be reassigned, just needs to be hm, hacked into the Bileto ;)09:05
sil2100I'll probably create a jenkins job for migrating entries, give me a few minutes09:06
sil2100Then it should be cool09:06
jibelsil2100, you can export the spreadshhet and recreate the request with the API09:08
sil2100Anyway, I'll get to it in a minute, just need to finish something up real quick first09:14
sil2100jibel: ok, on it09:23
sil2100jibel: looks like it won't be as easy09:46
greybacktrainguards: does the new dashboard show if silo ready for QA signoff? My silo11 should be in that state, but the dash doesn't show it09:58
jibelgreyback, it doesn't10:08
jibelgreyback, your request is in the old spreadsheet?10:09
greybackjibel: ah, that's still alive? Let me see10:09
jibelgreyback, status in the dashboard are not updated for request still in the spreadsheet10:10
greybackjibel: yep, ok spreadsheet reports it's ready for QA. Thanks!10:10
jibelsil2100, ^ data have to be migrated, we'll lose things otherwise10:11
sil2100jibel: yeah, I know, I tried migrating an entry as per robru's instructions but it doesn't work10:15
sil2100I'll try something10:16
sil2100ANyway, I have no idea why we couldn't have just left the spreadsheet working for the grace period10:17
sil2100Yeah, so it will be a bit more painful, the migration that is10:17
sil2100A watch-only rebuild is required10:17
sil2100(or actually any other operation)10:18
jibelsil2100, do you know how I differentiate deb package landings from tarball and clicks?10:37
jibelhow to*10:37
sil2100jibel: I think the only difference is if it has Manual Download Urls defined or not10:38
jibelokay, thanks10:38
jibelsil2100, where can I find the list of sources packages for a request? For example in https://requests.ci-train.ubuntu.com/v1/tickets?qa_signoff=Ready 'sources' is empty but it filled on the dashboard.10:44
jibelsorry for all the questions10:44
jibelsil2100, nm, found it10:46
sil2100jibel: k, no worries, just remember only robru knows the system well10:48
sil2100I'm a consumer here as well10:49
jibelsil2100, I know but he's sleeping and it helps asking questions. Look I found the answer ;)10:49
sil2100I'm a bit surprised with a few things, since after the public deployment so many things unexpectively changed10:49
sil2100jibel: btw. do you know if we still have HTTP access to the config files that were exported to lillypilly before?10:49
jibelsil2100, yeah it's related to my previous question10:50
jibelsil2100, https://requests.ci-train.ubuntu.com/static/json/index.txt will give you a list of the data available10:50
sil2100index.txt? No wonder https://requests.ci-train.ubuntu.com/static/json/ didn't work when I tried that10:51
jibelsil2100, then for the data itself about a silo replace index.txt by the silo you want10:51
jibelsil2100, https://requests.ci-train.ubuntu.com/static/json/ubuntu/landing-002 for example10:51
sil2100Yeah, got it10:51
jibelsil2100, fixing the bot is a bit more work than I expected TBH10:52
jibelsil2100, but now it should work for deb package landing, I need to validate click and tarballs now10:53
jibelbut it's a post-lunch task10:54
sil2100jibel: btw. migration might be really really painful to do10:56
sil2100Eh, now I feel bad I didn't have enough time to test bileto before deployment10:57
sil2100There's so many things I'd need to see fixed10:57
sil2100jibel: anyway, since one can't write a script that just migrates entries straight away as it requires 3 different isolated steps, I'm actually thinking of at least enabling the spreadsheet status updates11:00
sil2100Bileto silo assignment doesn't seem to work?11:04
sil2100robru: the assignment URL has some bad urlencoded signs, you can't assign any silos11:08
sil2100greyback_: hey! Could you try assigning your newly filled request?11:08
sil2100greyback_: I want to know if you also get the wrong urlencoded urls in the prepare job11:09
greyback_sil2100: ok11:09
sil2100Need to know if it's b0rken in overall, or just for me11:10
greyback_sil2100: looks broken: https://ci-train.ubuntu.com/job/prepare-silo/5561/11:10
sil2100greyback_: yeah, same here11:11
sil2100It copy-pastes correctly, but in reality it's wrong11:11
sil2100https%253A//code.launchpad.net <- things like these are in the request11:11
greyback_yeah, someone not urldecoding11:12
sil2100The funny thing is, I don't have access to hot-fix it11:12
greyback_funny ha ha?11:12
sil2100Very funny ;) I love having everything under IS control11:12
greyback_well there's no hurry with that silo, I need to wait until silo11 lands11:12
greyback_but can't say I'm impressed11:13
sil2100We can assign it by urldecoding by hand11:13
sil2100I'll try looking into fixing that11:18
davmor2jibel, sil2100: I'm trying to test silo 51 unfortunately I can't make it actually install anything it looks like there are just -dev and -doc files or am I missing something?11:20
jibeldavmor2, I don't even know which line of the spreadsheet it was11:22
davmor2jibel: 6011:22
jibeldavmor2, and the commit log is not helpful either "Initial revision for landing."11:23
mzanettijibel, hey, is there something wrong with silo 35?11:24
mzanettiI see lots of other silos coming in after that but getting tested beforehand11:24
davmor2jibel, sil2100: http://paste.ubuntu.com/11965973/11:24
davmor2jibel: citrain didn't pull anything in, I can't see a default package to actually install and apt refuse to do anything :)11:25
jibeldavmor2, no idea, ask michi for clarification11:25
davmor2jibel: will do11:25
jibeldavmor2, oSoMoN said silo 6 is ready but there is no card due to the death of the spreadsheet.11:26
bzoltan_sil2100: would you be able to help me with the silo assignement, please? I have just entered a new line to the new interface11:45
sil2100bzoltan_: as per e-mail, assignment is a bit broken right now - let me assign it for you with manual intervention11:46
sil2100I'm looking into fixing that, but I don't know robru's bileto code, so it'll take a while11:46
bzoltan_sil2100:  Yes, i have read the mail :)11:46
sil2100In the meantime I can manually help ;p11:46
bzoltan_sil2100: thank you11:46
sil2100greyback_: I'll assign your silo in a moment too, if you don't mind I'll use it shortly as test-bed11:48
greyback_sil2100: no prob, that's what I kept it there for11:48
sil2100greyback_: thanks11:53
=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
sil2100Strange, it looks like suddenly our CI Train jenkins simply stopped urldecoding the GET requests12:06
jibelsil2100, new bot updated and in production12:13
jibelubuntu-qa ^12:13
=== _salem is now known as salem_
sil2100jibel: sweet12:23
sil2100Ok, I still didn't find the reason why jenkins doesn't suddenly not urldecode12:23
* sil2100 goes of to lunch12:29
Laneysil2100: can we have a redirect for the dashboard please?12:32
Laneyrobru: ^ I guess :)12:33
jibelrobru, and also fix the link 'RAW' on the dashboard12:34
davmor2jibel: why are there 2 ticket for silo212:39
jibeldavmor2, the request ids in the spreadsheet and its replacement don't match12:40
jibeldavmor2, you can delete the second card12:40
davmor2jibel: they are breeding12:41
jibeldavmor2, it is just that the testing request exists in the spreadsheet and bileto. It has been created yesterday from the spreadsheet, then today I ported the bot to bileto which discovered this request. Since the request ids are different between the 2 systems it thought it was new and added a ticket12:48
jibeldavmor2, it will happen again if people create requests that already existed in the spreadsheet.12:49
davmor2jibel: it's the fact we are upto 3 that is concerning, it's like each time the import happens it creates a new ticket12:50
jibeldavmor2, yeah, I think I just fixed that12:50
pmcgowandavmor2, with click package updates in the custom tarball, do you know how these get applied during the update? is it via click install or is it via the image diff?12:50
davmor2pmcgowan: iirc the image pulls in the updates from trunk/store and then creates the tarball, then I believe it just uses image diff to install on the device, cwayne might be able to give you more clarity though12:54
cwaynethe ones in the custom tarball are via image diff i *think*12:56
jibelpmcgowan, it's a diff, everything is already unpacked in the tarball12:56
=== chihchun is now known as chihchun_afk
pmcgowanjibel, thats what I thought thanks, so there is no way a single click could fail to update13:01
bzoltan_sil2100: Do you know what should be the cure for the "error pushing:" failure of the ubuntu-device-flash?13:02
pmcgowanbzoltan_, is that arale? probably the flaky usb13:05
bzoltan_pmcgowan: no, it is both krillin and mako. Bootstrapping works fine, but simple flash fails13:06
pmcgowanbzoltan_, latest u-d-f from the PPA?13:07
bzoltan_pmcgowan: it should be up to date.. I check the versions13:08
pmcgowanI have  0.20-0ubuntu113:09
bzoltan_pmcgowan: I have 1.1+15.10.20150519-0~341~ubuntu15.04.1 from the recipe .. that is newer then any release13:12
bzoltan_pmcgowan: hmm... now it works... it could be USB flakiness13:13
bzoltan_pmcgowan:  it is random... "error pushing: error: protocol fault (no status)" and never finishes with the flashing :(13:20
infinityjibel: Who was your new victim for dkms testing issues?13:21
infinityjibel: Looks like the current builds have been running for 27 hours. :P13:21
jibelinfinity, still me13:21
pmcgowanbzoltan_, that weird maybe your laptop13:22
infinityjibel: Oh.  Well, hi you.  Iz broke.13:22
bzoltan_pmcgowan:  it happens on both my laptops... checking the third one13:22
pmcgowanbzoltan_, I am not sure where you got that version13:23
pmcgowanI don't see that in the ppa13:23
bzoltan_pmcgowan: from here -> https://code.launchpad.net/~bzoltan/+recipe/sdk-tools-dev13:23
pmcgowanbzoltan_, why not the one in the ppa?13:24
jibelinfinity, the execution node is down again13:24
jibelinfinity, I'll restart it13:24
bzoltan_pmcgowan:  :) somebody need to test the release candidate of the phablet-tools.13:25
bzoltan_pmcgowan:  but i will  downgrade now and see... if it works then the phablet-tools trunk is broken13:25
infinityjibel: Is that same node used by autopkgtests, cause they seem a bit backed up too.13:25
jibelinfinity, yeah but adt jobs are distributed, it shouldn't be a problem13:28
jibelinfinity, let me restart whatI can13:28
=== chihchun_afk is now known as chihchun
jibelinfinity, dkms node is back13:34
infinityjibel: Will the jobs need a kick, or should they wake up?13:34
jibelinfinity, they'll wake up13:34
infinityjibel: Ta.13:34
jibelcihelp: ^ can you monitor wazn-adt, it's the second time it crashes in 2 weeks and no one notice13:35
fgintherjibel, it's just the 'wazn-adt' slave right. not the node itself or the two other slaves running there?13:38
jibelfginther, yes, the jenkins-slave process dies and doesn't respawn13:38
boikosil2100: does the dashboard still show the silos as marked as tested and ready for QA?13:41
fgintherjibel, I'm a little curious as to what's going on here. initctl reports 'auto-package-testing-jenkins-slave stop/waiting', but the slave is running. Do you know how it was restarted?13:42
sil2100boiko: yes, but not from the spreadsheet... sadly Robert removed support for the spreadsheet completely13:42
boikosil2100: so, yesterday I marked some silos as tested on the spreadsheet, was that migrated or do I need to mark them as tested again?13:43
sil2100boiko: no, all is fine13:43
boikosil2100: I just don't see them marked as so in the dashboard13:43
sil2100boiko: yes, since as I said, the dashboard doesn't understand the spreadsheet anymore, and there was no migration to bileto yet13:44
sil2100So it's just marked on the spreadsheet, QA is aware of those13:44
boikosil2100: ok, good, just checking, thanks for the info :)13:44
sil2100This switch to bileto is a bit faulty13:45
fgintherjibel, Oh, I think I see now. Looks like you manually started it.13:46
sil2100fginther: hey! Did you have an issue with jenkins and the parambuild plugin before that suddenly it wasn't urldecoding the parameters from the URL?13:51
fginthersil2100, no, I can't say I've seen that13:52
jibelfginther, sorry was otp. I restarted it13:54
fgintherjibel, thanks for clarifying13:54
infinityjibel: I question the efficacy of your restart.  Or your claim that the jobs would unstick themselves. :)13:56
infinityjibel: All of them still seem to be in the "building for over a day" state, I'd have expected a few to be done by now.13:56
cyphermoxhi, could I please have a silo for line 44, or should I open a bileto ticket now instead of using the spreadsheet (line was already there, been sitting in the queue for over a week)?13:57
sil2100cyphermox: please open a bileto ticket :)13:57
cyphermoxsil2100: thanks!13:57
greyback_jibel: hey did the current silo11 slip through the cracks? I don't see it in the QA testing board14:01
cyphermoxsil2100: am I also the person supposed to hit assign?14:01
sil2100cyphermox: yes, but it's broken a bit right now, let me assign it for you14:02
cyphermoxah ok14:02
cyphermoxjust wanted to know if it was all self-serve ;P14:02
cyphermoxawe_: ^^14:02
cyphermoxnext time, you'll be able to do it all yourself14:02
cyphermoxmorphis: ^ I guess this is relevant for you too14:03
sil2100Aaah, I see the issue14:03
sil2100Not sure why suddenly firefox decides to urlencode already urlencoded strings14:06
jibelinfinity, jenkins is slowly digesting the queue. The UI is a bit confusing but if you scroll down to a node called wazn-adt, you'll see that all the builds with a number in front of them are recent.14:08
jibelinfinity, the jobs without number are virtual jobs that jenkins create to keep track of its upcoming tasks14:09
rvrboiko: ping14:09
rvrsil2100: To approve a silo, is the spreadsheet still working?14:10
sil2100Looks like an apache misconfiguration14:10
sil2100rvr: yeah, for approving yes - since those are simple formulas14:10
infinityjibel: Ahh, check.  I'll attempt more patience.14:10
jibelgreyback_, it probably slipped through the cracks14:11
rvrsil2100: Ok14:11
greyback_jibel: shall I add a card?14:11
jibelgreyback_, which line of the spreadsheet is it?14:12
boikorvr: pong14:12
greyback_jibel: 5914:12
greyback_jibel: thank you14:13
jibelgreyback_, added14:13
rvrboiko: Silo 814:13
boikorvr: yep14:14
rvrboiko: Does the bug has any manual test?14:14
rvrboiko: Skip network checks on always dispatchable accounts14:14
boikorvr: ah yes, let me check the testplan, salem_ said something about addind that to the testplan, let me double check14:15
boikorvr: the manual test is: disable data connections (both wifi and mobile data) and reboot the device14:15
boikorvr: the dialer needs to show the carrier name, if it doesn't, that's the bug14:16
salem_rvr, yes, just ensure there are not network (wifi/data) connections active after a reboot and try making/receive calls/messages.14:16
rvrsalem_: Ok, I'll check that. Please update the test plan.14:17
salem_rvr, will do, thanks.14:17
sil2100robru: I see that you're double urlencoding all your requests - it was done like this in the spreadsheet as well14:26
sil2100robru: now apache seems confused by this and triple-urlencodes everything14:27
sil2100robru: is there any reason we have to urlencode urlencoded strings?14:27
davmor2sil2100: silo006 is good to go14:31
salem_rvr, just updated telepathy-ofono testplan.14:37
rvrsalem_: Hmmm14:40
rvrSomething weird14:41
rvrI'm in a call, but dialer app doesn't show any call14:41
bzoltan_sil2100:  how in an automatic test environment can I unlock the screen?  Once I have adb connection :) I could unlock it with a gdbus call... but there is no adb when the screen is locked14:41
salem_rvr, is the green call indicator at the top?14:41
rvrsalem_: Nope14:41
salem_rvr, incoming call?14:42
rvrsalem_: Yes14:42
rvrsalem_: Second try seems good14:42
salem_rvr, do you have wifi/data connection?14:43
rvrsalem_: Nope, deactivated them14:43
salem_rvr, ok, but it was like this when you got the incoming call, right?14:43
pmcgowanhey udf is tellig me --revision is unknown flag, what am I doing wrong14:44
rvrsalem_: Yes. call was established but then the UI reverted to the offline state, even when the call was active. I could talk, etc.14:44
salem_rvr, that's really weird, but I would say that's a separate bug.14:45
sil2100Damn, this bug is a waste of time14:45
sil2100robru: anyway, found how it's happening that it's broken, but I need you to understand why it's happening14:46
sil2100bzoltan_: I think once developer mode is enabled then the screen doesn't need to be unlocked to get adb access14:47
sil2100bzoltan_: at least that's how my device is working ;)14:47
salem_rvr, the original bug would prevent you from receiving the call in first place.14:47
pmcgowanpopey, davmor2 ?  hey udf is tellig me --revision is unknown flag, what am I doing wrong14:47
bzoltan_sil2100: makor or krillin? I have flashed with --developer-mode --password=${PASSWORD}14:48
sil2100bzoltan_: I'm using arale here with developer mode, but I also switched the switch in system-settings14:48
davmor2pmcgowan: --revision has to go before touch in the command14:48
davmor2pmcgowan: should be u-d-f --revision x touch --channel whatever/channel14:49
pmcgowandavmor2, thanks so lame14:49
popeyi love u-d-f14:49
jibelwhat's the logic to have this option before touch14:49
popeyI have to re-learn it every time14:49
jibelevery one try after14:49
bzoltan_sil2100:  I hve developer mode enabled and the adb gets closed when the lock screen comes up ... on mako with vivid overlay14:49
pmcgowanyeah there is even an open bug14:49
sil2100bzoltan_: hmm, that's strange - could you poke cihelp about that? They might know more since they use devices in the lab for testing14:50
sil2100So they should know how to properly unlock it14:50
jibelbut u-d-f does everything, creates images, flash them, query the server...14:52
pmcgowanit just needs to parse arguments now14:52
jibelah right, it's a detail14:54
fgintherbzoltan_, sil2100, the krillins and arales require a special recovery image to be used during u-d-f flashing14:55
fgintherwhich enables adb while the screen is locked14:55
bzoltan_fginther:  I amnot sure if I understand that14:55
bzoltan_fginther:  I am on mako this time ..14:55
fgintherbzoltan_, oh, mako doesn't require it14:55
davmor2pmcgowan: run u-d-f --help and it lists the order of arguments and options as you build up the command14:56
fgintherunless things have changed very recently14:56
bzoltan_fginther:  my mako with the latest vivid image it does14:56
bzoltan_fginther:  I see in the --help it says about the --recovery-image= so, where can i get that?14:57
jibeldavmor2, RTFM is not a good answer :) when a user does the same mistake each time he uses a tool, it supposes something is wrong with the tool14:57
rvrsalem_: boiko: Silo 8 approved14:57
fgintherbzoltan_, http://people.canonical.com/~plars/touch14:58
salem_rvr, awesome, thanks14:58
bzoltan_fginther:  Thank you ... I will ask plars to make a mako version too >(14:58
davmor2jibel: no that was a genuine piece of advice,  just append --help and it fills you in on the options available after each step not saying that it is a good thing just a helpful one :)14:59
fgintherbzoltan_, these were not made by plars, that was just the only easily place we had to store them14:59
pmcgowandavmor2, write a qml front end to it please :)15:00
fgintherbzoltan_, let me see if I can dig up who made them, but if mako now needs them :-(15:00
plarsbzoltan_: oh, we need one for mako now also? I just grabbed those from an older image because there was no official place at the time, there may be some official place now though15:00
plarsbzoltan_: maybe sil2100 knows?15:00
davmor2pmcgowan: if you see  the code I write you wouldn't even joke about that ;)  There are way more competent coders in the room ;)15:00
bzoltan_plars: Yes, I was strugling my way thru the tools today to figure it out15:01
fgintherbzoltan_, FWIW, I can see version 202 from ubuntu-touch/rc-proposed/ubuntu channel being flashed on mako with no need for a recovery image15:03
robrusil2100: as a workaround for right now, it works if you s/%2520/ /g on the text field and leave the URLs doubly encoded: https://ci-train.ubuntu.com/job/prepare-silo/5565/console15:09
jibelrobru, sil2100 when I log into https://requests.ci-train.ubuntu.com/ I just get the header and no request15:18
jibelI checked "ci-train-users" on the sso page15:18
robrujibel: yeah I was getting an error as well, it seems intermittent, try reloading a few times15:19
sil2100Same here15:19
sil2100Yeah, after a few reloads it works now15:19
jibelah it is just terribly slow15:19
jibelrobru, you didn't have to reimplement everything from the spreadsheet :P15:20
robrujibel: no no, what's happening is that it's doing an ajax request in the background that silently fails, it retries every few minutes. it's fast when it works15:20
robrujibel: and of course I can't access the log, will need IS to poke at it, sigh15:21
jibelrobru, where should I file bugs? when I edit then cancel it logs me out15:22
robrujibel: yeah file bugs against lp:bileto, but again this is the same issue, ajax request is failing intermittently15:22
jibelI'll be patient until you get the logs then15:23
boikorvr: thanks15:24
LaneyI just put a redirect in place for the old dashboard URL16:07
robru_sil2100: jibel: ok sorry my IRC client has picked the perfect time to explode16:11
robru_sil2100: I'm in webops trying to get things sorted out, apparently the problem is logrotate misconfigured and then bileto can't log properly and gets angry16:12
robru_sil2100: jibel: it should be temporarily ok now, but I'm working on a permanent fix16:12
Laneyis the spreadsheet still around enough to let me reconfigure a silo?16:13
robru_Laney: yes16:15
robru_Laney: the spreadsheet won't update status but it can reconfigure an existing silo16:15
LaneyOK, thanks!16:16
=== chihchun is now known as chihchun_afk
Laneyrobru_: do I need to worry about the messed up urlencoding in MERGE_PROPOSALS and SOURCES?16:29
robru_Laney: yeah, you should manually s/%2520/ /g to make it work (the urls can stay double encoded but you need real spaces between the urls)16:31
dobeythat's an odd set of packages in a silo16:48
robrusil2100: what were you saying about SSO redirecting through http? I'm not seeing that16:55
sil2100robru: let me check if it's still doing that16:55
sil2100robru: yeah, still does that, let me PM you with details16:56
=== alan_g is now known as alan_g|EOD
robrucihelp: can I get CI configured for lp:bileto the same way it already is for lp:cupstream2distro? thanks17:42
psivaarobru: looking into it17:44
robrupsivaa: thanks17:45
jibelrobru, when I can the API, I frequently get an 'Internal Server error'17:49
jibelI call*17:49
robrujibel: hmmm I thought I fixed that17:50
jibelrobru, this is what I call: https://requests.ci-train.ubuntu.com/v1/tickets?qa_signoff=Ready17:50
jibelrobru, actually I get it more often than the valid output17:50
robrujibel: yeah I'm seeing that too, even without the query17:51
kgunnjibel: hey since you're on :)17:52
robrujibel: oh well this time I'm finally seeing a traceback in the logs17:52
kgunnthat one can be unblocked now17:52
robrujibel: ok I don't know when I can fix this by but at least I have something to go on now17:52
kgunnand that was our goof, sorry17:52
jibelkgunn, done17:53
jibelrobru, ok17:54
boikojibel: hi, in the spreadsheet I had silos 15 and 40 marked as ready for QA to test, but I don't see it in the QA spreadsheet, could you please just check why they are not there?17:56
boikorobru: quick question: rvr has tested and approved silo 8, do I need to do anything for it to get published?17:57
robruboiko: pinging me is a good start17:57
boikorobru: still getting familiar with the new UI, glad we got rid of the spreadsheet :)17:58
robruboiko: yeah things are exploding right now, sorry this transition is going to be a little rough17:59
boikorobru: no worries,I will just be pinging you more often but no worries regarding transition problems, that's fine18:01
AlbertAcihelp: ^ this is powerd so it will fail boottest due to the package install failing to update config-default.xml being in a seperate ro partition18:03
jibelboiko, they are not there because the spreadsheet doesn't update the statuses in the dashboard anymore because of the migration to the new system.18:04
jibelboiko, we'll add the cards18:05
jhodapprobru, is the second to last line about Bileto going to be fatal or just ignore? https://ci-train.ubuntu.com/job/ubuntu-landing-038-1-build/42/console18:07
robrujhodapp: this is a spreadsheet landing right? if so it's fine, it's just trying to push the status to bileto and bileto doesn't recognize the id number. not fatal18:07
robrujhodapp: if that's a request from bileto though then that's a problem (still not fatal but not good)18:08
jhodapprobru, yeah it was a spreadsheet landing18:08
robrujhodapp: yeah no worries then18:08
jhodappexcellent, thanks!18:08
robrujhodapp: you're welcome18:08
jhodapplooking forward to using your work in bileto18:09
jibelboiko, what are the lines for silos 15 and 40?18:11
boikojibel: let me look, just a sec18:13
jibelboiko, nm, found it18:13
boikojibel: ok18:13
jibelrobru, no pressure, but hurry up, I won't create cards manually for a long time :)18:13
robrujibel: yeah sorry. I will migrate the spreadsheet later today, just putting out fires right now18:15
fgintherAlbertA, thanks for the heads up. Looks like it hasn't hit proposed just yet as it isn't in the excuses report18:22
psivaarobru: http://s-jenkins.ubuntu-ci:8080/job/bileto-ci has been added but http://s-jenkins.ubuntu-ci:8080/job/bileto-ci/1/console sees some failures. If you'd please take a look and let us know what are the dependencies, we'd be able to fulfill18:48
robrupsivaa: oh thanks18:49
robrupsivaa: one sec18:49
robrupsivaa: the dependencies are here: http://bazaar.launchpad.net/~cupstream2distro-maintainers/bileto/charm/view/head:/hooks/install#L1918:49
robrupsivaa: of which some are only accessible in trusty-cat-ue archive18:50
robrupsivaa: you'll need to add "deb http://archive.admin.canonical.com/ubuntu trusty-cat-ue main" archive to get all the deps18:51
psivaarobru: ack, will take a look at adding them18:51
kgunnrobru: hey, so i just tried to reconfig silo 0 via the old sheet....is that doable ?18:52
kgunne.g. for old silos, do we still go thru the sheet?18:52
robrukgunn: yes the spreadsheet should be usable for reconfiguring, what are you seeing?18:52
kgunnrobru: it could very well be me18:52
robrukgunn: yes for this moment we go through the old sheet for reconfiguring existing silos. I'm working on migrating the data but it's not ready yet18:52
robrukgunn: yeah, it's you: https://code.launchpad.net/~mir-team/mir/silo0/+merge/264015 40418:53
kgunnrobru: nvmd (sheepishly)18:54
kgunnrobru: helps if you actually paste the link you want in the cell....18:54
robrukgunn: lol no worries. I should clean up those tracebacks and make the errors more readable18:54
robrupsivaa: certain failures are expected at this point, if you get a CI run that looks like this: https://pastebin.canonical.com/136541/ then you know all the deps are available and everything is working18:56
kenvandinetrainguards: silo 2 is marked as QA granted, but the status on the dashboard doesn't say it's ready to publish19:10
robrukenvandine: yeah sorry it's a mess, thanks for pinging me19:10
kenvandinerobru, no worries, just want to make sure nothing's holding it up :)19:11
* kenvandine really wants to see passing CI again :)19:11
robrukenvandine: well, no statuses are being reported anywhere, so that's holding it up19:11
robrukenvandine: if you see it has QA granted (in the trello board?) you can publish it19:11
kenvandinei'll do that :)19:11
kenvandinerobru, just making sure19:11
robrukenvandine: dashboard won't report qa statuses for spreadsheet stuff19:11
robrukenvandine: you're welcome19:11
jibelrobru, brendand marked is qa granted in bileto19:12
kenvandineyeah, that's where i saw it19:12
kenvandinenot the spreadsheet19:12
kenvandineugh... should i be concerned with this?19:13
kenvandine2015-07-30 19:12:49,226 ERROR Bileto says: 400 BAD REQUEST19:13
kenvandinebut it finished with SUCCESS19:13
davmor2robru, sil2100: silo51 is good to go19:13
robrudavmor2: thanks19:13
AlbertAfginther: powerd is listed in excuses now.19:21
kenvandinerobru, i think something blew up there publishing to wily, silo 219:24
kenvandinei see it in the overlay ppa for vivid, but no change mail yet for wily19:24
kenvandinethat 400 error makes me think something didn't happen19:25
robrukenvandine: no that's fine19:25
robrukenvandine: the 400 error is just bileto saying "I don't know this requestid"19:25
kenvandineit's been over 10 minutes and no wily change mail19:25
kenvandineit's usually quicker than this19:26
robrukenvandine: log?19:26
boikorobru: silo 34 was saying address-book-service is outdated, so I tried rebuilding it, but it failed with the error seen on the dashboard19:27
kenvandinerobru, maybe it will work... just i got worried :)19:27
robrukenvandine: that log is fine but indeed cicopy is broken19:27
boikorobru: there was actually no landing in there, but the changelog shows an entry from the citrain: https://launchpad.net/ubuntu/wily/+source/address-book-service/+changelog19:27
robruboiko: ok just wait a moment19:27
kenvandinerobru, ok, so you'll handle it?19:28
boikorobru: no hurry, just letting you know19:28
robrukenvandine: yeah19:28
kenvandinerobru, thx19:28
robrukenvandine: ok it should be fixed, but the cicopy thing only runs every 5 minutes, so you should see that within 5 minutes or so. I gotta eat, brb19:33
robrukenvandine: ok looks good19:36
robruboiko: yeah so address-book-service is in -proposed, I'm assuming another silo already published and hasn't merged yet. that's why you're getting that error.19:37
robruboiko: in fact silo 28 needs to migrate before you can rebuild19:37
boikorobru: ah ok, found it, silo 2819:38
boikorobru: ok, thanks19:38
robruboiko: you're welcome19:38
kenvandinerobru, thx!19:41
kgunnrobru: can you help me with a game of what-am-i-doing-wrong19:42
robrukgunn: in a sec, i have some food on the stove19:42
kgunnrobru: nw, i'll type here, just when you get a chance19:43
kgunnso in silo42, it gave me an error about changelog versioning, so i bumped it19:44
kgunnit's just a demo silo, so i changed the deb version string from 2015 to 2016, thinking that should make it not complain for a while19:45
kgunnbut what's weird is in the output19:46
kgunndon't know how why it's comparing to another deb string containing 2016 (as in the year)19:47
robrukgunn: your using a branch that has been released to wily but your silo targets vivid. If you read the error not closely it says 15.04 is less than 15.10.19:54
robrukgunn: you need to either do a dual silo or change your changelog to say 15.0419:54
kgunnah ok19:54
robrukgunn: your welcome19:55
kgunnrobru: when specifying dual, do we also need to specify destination ppa (stable phone) ?20:02
robrukgunn: no don't do that20:02
robrukgunn: dual is always overlay. If you set the overlay that means you want wily packages in the overlay20:03
kgunni didn't do that...i saw others by convention, just confirming20:03
robrukgunn: i may have put in a check to stop that from happening but i can't remember if I just dreamed it20:04
kgunni think it was always the same in the old sheet too20:05
kgunne.g. possible, but no one used it20:05
fgintherAlbertA, powerd should be unblocked shortly20:05
robrukgunn: dual silo is just a wily silo with a vivid overlay tumor, all the options control the wily half, vivid half is all hard coded20:05
kgunnrobru: ^ just a little bug, i switched to dual, then hit "assign" and i thot huh, why does it say vivid....then it failed ^, second time it says "dual" like i expected20:08
AlbertAfginther: ack20:13
robrukgunn: maybe your clicked assign before the edit saved?20:14
robrukgunn: there's an ajax roundtrip after clicking save, should be fast but maybe it was slow that time or something20:14
kgunnrobru: i don't think so...but i do feel like i'm getting logged out incessantly20:14
robrukgunn: yeah the log out thing is just the backend being crap, I'm working on it20:15
kgunnthat was prolly it20:15
kgunnjust raced out20:15
AlbertArobru: how do I sync packages from wily to vivid+overlay now? Do I put ubuntu,wily in Sync source? and package names in "Manual Source Packages" box?20:27
robruAlbertA: yeah20:37
robruAlbertA: hover over the fields should explain a bit, please email me if anything is unclear and i can expand it.20:37
AlbertArobru: thanks yeah they were self-explaining...just wanted to confirm :)20:38
AlbertArobru: could we have a ubuntu,wily option in the dropdown in Sync source?20:38
robruAlbertA: also if you double click on the sync field it suggests some common options20:39
robruAlbertA: Hmmmmmmm if that not in there already?20:39
AlbertArobru: I don't see it in the dropdown20:44
robruAlbertA: ah I see it's just ubuntu,vivid. Ok I can add that but I won't be able to roll it out until later20:44
AlbertArobru: also can I delete duplicate entries...I got logged out  while saving one clicked again.20:45
robruAlbertA: you can't delete them but just set the status to 'Abandoned' and it'll disappear20:45
AlbertArobru: cool thanks20:45
robruAlbertA: you're welcome20:45
=== chihchunl is now known as chihchun
=== cwayne- is now known as cwayne
charlesis anyone else having trouble staying logged into https://requests.ci-train.ubuntu.com/ ?21:11
dobeyi can't log in at all on it; it just always has the "Log in with Ubuntu SSO" button21:13
dobeytrainguards: ^^21:13
robrudobey: charles: yes there are some known issues unfortunately, keep trying21:14
robrucharles: dobey: if you get logged out, don't actually try to log in again, just reload the page.21:14
charlesrobru, ack21:15
charlesthanks for that; I've been re-logging in several times21:15
dobeyoh now there's some content21:15
charlesya, same here21:15
robrucharles: dobey: yeah sorry I need to make it give a real error message, right now unfortunately it just appears logged out without actually being logged out21:15
dobeywell i can't see how to actually create a new request21:15
robrudobey: you need to log in to get the new request form21:16
charlesrobru, no worries, rollouts are never perfect. I appreciate the tip21:16
robrudobey: but if it's not showing any requests at all, that's a backend issue, requests should be visible even if logged out21:16
dobeyi thought i was logged in21:16
robrudobey: if you can see the requests but still the login button then you're definitely not logged in21:16
robrudobey: if you can't see the requests but see the login button, that's an error state that's indeterminate.21:17
dobeyrobru: i mean, i clicked on log in and performed that loop about 20 times21:17
robrudobey: what are you seeing?21:17
dobeywell now i can see the request form21:17
dobeyso apparently i wasn't logged in21:17
charlesdobey, that's what I was seeing as well. I'm not sure what changed s.t. I finally got logged in21:18
robrudobey: it's possible. I don't know what the hell is happening, stuff that worked fine in staging is exploding in production.21:18
dobeybut seems backend issues make it difficult to know whether i am actually logged in or not21:18
dobeyrobru: that's the problem; you put it in production :)21:18
robrucharles: dobey: a good way to troubleshoot is to visit https://requests.ci-train.ubuntu.com/v1/tickets. If you get a 500 error, the backend is broked, try reloading. if you get a json blob then it's all good21:18
dobeycould use some "responsive design" improvements; i like small fonts, but with these colors, it's a bit too small to fit in a single window without horizontal scrolling on my crazy 4k monitor :)21:21
dobeyah well21:21
robrudobey: what? I have a 1080p that doesn't horizontally scroll21:21
dobeyrobru: doesn't horizontally scroll with a full screen/maximized browser you mean?21:22
robrudobey: yeah, maximized21:22
dobeyyeah, i use windows21:23
robrudobey: well, considering the spreadsheet would scroll horizontally through three screenfuls I think this is a win. if you can't maximize your screen then I can't help you :-P21:23
dobeybut my browser is 1266px wide21:24
robrudobey: why is it even horizontally scrolling? shouldn't the longer cells be wrapped?21:24
dobeyrobru: eh, i hated the spreadsheet horizontal scrolling too. but it's a spreadsheet, and it's google's, so not much we can do about that21:24
robrudobey: somebody already complained about horizontal scrolling and I already fixed it21:24
dobeysome are wrapped21:24
dobeybut the new request block at the top isn't wrapping all the text entries21:25
robrudobey: it's a free form text field, why would it wrap?21:25
dobeyrobru: why do i need four of them in the same row? why can't they stack?21:26
robrudobey: they correspond to the displayed rows beneath. it's like a row header.21:26
dobeyso why can't the stuff below wrap at the same places too?21:28
robrudobey: because it's tabular data inspired by the spreadsheet. I basically took one huge google row and broke it into two manageable rows.21:28
charlesrobru, it doesn't stop work from getting done, but minor bug report: Status for the pay-service req I'm on now is listed as "Silo ready to build" but it started building 7 minutes ago21:32
robrucharles: yeah that's probably related to the 500s in the backend21:32
robruAlbertA: 11 is ready for qa?21:58
AlbertArobru: silo 8 yes...21:59
michirobru: Can you help with this? https://ci-train.ubuntu.com/job/ubuntu-landing-051-2-publish/6/console21:59
robrumichi: one sec21:59
robrumichi: you need an archive admin, please ask in #ubuntu-release. I already pinged there and got no response22:00
michiAh, I see. Thanks, will try!22:01
charlesrobru, so if I don't need QA signoff and am ready to publish, I would do that by editing the signoff status, yes?22:05
robrucharles: yes you can set that to 'Publish without QA'22:05
charlesrobru, thanks22:05
AlbertAtrainguards: I don't see a card entry in the QA trello board, for silo 8, which is ready for QA... do I just need to wait?22:31
robruAlbertA: no you need to ping qa if it's a spreadsheet landing, everything is broken I'm afraid.22:32
AlbertArobru: it's an entry in the new one22:32
robruAlbertA: what number?22:33
robruAlbertA: ok, bileto is literally on fire, so qa bot probably can't access the API to know that it's ready. ping some qa person to get it on the queue22:33
AlbertArobru: ack thanks22:34
thomirobru: umm "literally on fire"22:34
robruthomi: LITERALLY22:34
thomi...pics, or it didn't happen :P22:34
robruthomi: http://images.rapgenius.com/4jh2iqo0mlekbviwkr97w46m6.450x600x1.gif22:35
=== salem_ is now known as _salem
AlbertArobru: looks like duplicate entries to the QA trello board are being added periodically...I saw 3 about 15mins later 4, about another 15 mins later 5...https://trello.com/b/AE3swczu/23:40
AlbertArobru: ooh now 6....23:40
robruAlbertA: that would be qa's job to fix, I don't make those23:40
AlbertArobru: ack23:41
robruAlbertA: most likely jibel's bot has a bug23:41

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