[00:57] if I change one of the MPs I'm wanting to land, do I need to do anything for the silo to pick i tup? [00:58] * Chipaca is rebuilding but it picked up the old MP [01:09] looks like 'reconfigure' is what i wanted === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [01:56] hrm. that didn't work. [02:04] ok, so reconfigure definitely does pick up the right branch [02:04] === trainguard: IMAGE 185 building (started: 20140812 02:05) === [02:27] * Chipaca tests [03:08] psivaa: I still got a build failure for the same reason - http://s-jenkins.ubuntu-ci:8080/job/ubuntu-sdk-team-ubuntu-ui-toolkit-staging-utopic-amd64-ci/607/console or was it started before the fix landed? [03:29] is anybody around who could cihelp? [03:30] bzoltan, hello [03:30] bzoltan, looking [03:32] fginther: there was a gcovr fix/update on utopic to solve that problem. I do not know what time exactly and I do not know what timezone Jenkins is in. [03:32] bzoltan, the gcovr fix was introduced in 3.1-1ubuntu1, which first appears in build 608 [03:32] the full build log should always show you the version of all packages used [03:32] fginther: but looking at the http://s-jenkins.ubuntu-ci:8080/job/ubuntu-sdk-team-ubuntu-ui-toolkit-staging-ci/ it seems that #773 still has that problem [03:34] slangasek: ohh.. that is true, thanks [03:34] bzoltan, correct, the build numbers don't necessarily sync up. The 608 I referred to was specifically for http://s-jenkins.ubuntu-ci:8080/job/ubuntu-sdk-team-ubuntu-ui-toolkit-staging-utopic-amd64-ci/608/ [03:34] fginther: but this is just one isse... the other issue is why UITK builds take 5-6 hours? [03:36] the build times seem to have spiked sometime around build 734, what changed then? [03:36] bzoltan, there has been an excessively high load on jenkins today. Part due to the gcovr issue [03:36] fginther: and the third question is, why so few MRs are picked to be build? During the last ~10 hours only few were picked [03:36] bzoltan, load is still a problem: http://s-jenkins.ubuntu-ci:8080/computer/(master)/load-statistics?type=min [03:36] lots of jobs are in the queue still [03:37] fginther: but the build time is the actual time the job is running, no? [03:37] i.e., queue times are bad but should be unrelated to http://s-jenkins.ubuntu-ci:8080/job/ubuntu-sdk-team-ubuntu-ui-toolkit-staging-ci/buildTimeTrend [03:37] fginther: I understand, but we could not get a single UITK build in the last 4-5 days. Not a single one... [03:37] bzoltan, ok, that is bad [03:38] fginther: it is very very bad.. because we can not respond to real failures and our MRs are just piling up and so we risk of new conflicts and new problems [03:38] slangasek, ubuntu-sdk-team-ubuntu-ui-toolkit-staging-ci is a 'parent' job. It triggers a number of jobs, some of which are stuck on the build queue [03:39] ah, ok [03:39] slangasek: fginther: i sthere anything I should/could do? I am considering to skip using Jenkins. [03:39] bzoltan, are these missing MRs related to lp:~ubuntu-sdk-team/ubuntu-ui-toolkit/staging ? [03:39] === trainguard: IMAGE 185 DONE (finished: 20140812 03:40) === [03:39] === changelog: http://people.canonical.com/~ogra/touch-image-stats/185.changes === [03:40] bzoltan, or a different branch? [03:40] fginther: al our MRs are targeting the staging [03:40] fginther: that is the branch we stage our development and I release that staging to the archive and back to the trunk occasionally ... like once a week. [03:45] slangasek: fginther: But I conducted the last big landing already with manual merging to this staging branch. Jenkins landed less then 20 MRs in the last 15 days... I have 25 active MRs in my queue. It is horror... with the present trend I expect 2-3 weeks just to land on our staging. This is a serious problem for us. [03:51] bzoltan, I'm looking through a few jobs to better understand what's going on here. The gcovr issue impacted lots of MRs, but it's not the only 'oddity'. One thing we can try to do is scale back a bit, like remove coverage from the staging builds [03:51] bzoltan: I don't understand what this job is or what the success criteria are; are you expecting all of these subtests to pass? [03:52] slangasek, 'all pass' is the expectation [03:52] hmm [03:52] seems to be quite far from that currently [03:52] slangasek: fginther: scaling down sounds good to me... [03:53] oh, I was reading things wrong (upstream projects vs. downstream projects) [03:53] ok [03:54] slangasek: fginther: I am ranting pretty much without any success that oversecuring quality with super heavy builds ended up so much wasted time, that if we add up the lost time we could fix dozens of regressions. 5-6 hours for a UITK build what takes few minutes? 10 MRs and we lost a week of an engineers. Sounds expensive to me. [03:55] bzoltan: well, er, who is responsible for this job configuration, if not you and your team? I don't know anything about this setup, I don't think ranting at me will be any more successful :) [03:56] bzoltan, there's no reason not to explore options. The coverage build for a staging build is not all that useful (when it also takes place elsewhere) [03:56] slangasek: Sorry, mate :) it s certainly not you who should do anything :) with this. [03:57] slangasek: these configurations were extended by time... I have not added any extra crap on them ever. [03:57] right... but who owns them? [03:57] fginther: can we switch off the coverage builds? [03:57] http://s-jenkins.ubuntu-ci:8080/job/generic-mediumtests-builder-utopic-armhf/4612/console fun [03:57] morning [03:58] slangasek: "somebody" (tm) in some otherteam [03:58] bzoltan, yes. I can get to that before going to bed [03:58] that seems suboptimal ;) [03:58] fginther: thanks a lot! [03:59] slangasek, bzoltan, ownership was a lot simpler when CI and QA were the same team [03:59] now it's a very gray area [03:59] this one doesn't appear to have anything to do with gcovr. http://s-jenkins.ubuntu-ci:8080/job/ubuntu-sdk-team-ubuntu-ui-toolkit-staging-utopic-armhf-ci/611/console [04:00] fginther: is it? I would have assumed that QA + the upstream team are responsible for defining what tests should be run [04:00] slangasek: that is an API test failure. The SDK team is to blame for that. [04:01] slangasek, that is the basic idea, although it is constrained by what CI is capable of doing [04:01] bzoltan: ok - so you've seen it and understood it, fine, I'll leave it in your hands then :) [04:01] slangasek: the optimal would be to have a full time integrator for the bigger and more critical assets. [04:03] slangasek: the UITK is responsible for many failures, but hen the turnaround is 1-2 days for a new MR then responding to failures is difficult. [04:03] sure [04:03] slangasek: sometimes feels like 80's ... you take your code cards to the univ and go back next week to check the results :) [04:04] slangasek, big thanks for pointing out that mkdirs failure. That's a corrupted node [04:05] so http://s-jenkins.ubuntu-ci:8080/job/generic-deb-autopilot-utopic-touch/buildTimeTrend seems to be the job that's the limiting factor here? [04:05] or is that just a wrapper around http://s-jenkins.ubuntu-ci:8080/job/generic-deb-autopilot-runner-mako/ + http://s-jenkins.ubuntu-ci:8080/job/generic-mediumtests-builder-utopic-armhf/ ? [04:09] slangasek, yep, it's essentially a wrapper [04:10] http://s-jenkins.ubuntu-ci:8080/job/generic-deb-autopilot-runner-mako/3290/console takes a while, but it's not obvious from the log where it's spending its time... the only bits with timestamps are all within a couple minutes of each other [04:11] fginther, slangasek, bzoltan: Hey guys, looking at the convo, it looks like the autopilot runs are taking longer than expected? [04:12] veebers: I don't have enough info to draw that conclusion [04:13] could also be setup/teardown stuff taking long [04:13] http://s-jenkins.ubuntu-ci:8080/job/generic-deb-autopilot-runner-mako/3290/console seems to show 10 minutes before autopilot is ever called [04:14] slangasek: ok, is there anything I can do to help currently? I'm interested if it is autopilot itself. [04:14] for instance, is there a point in time where these job times go up? I would like to see if it matches the most recent release of AP [04:14] and only 2 minutes from the start of the autopilot run to the last output from that job, AFAICT [04:15] so I don't think autopilot itself was the bottleneck here; 2 minutes of tests, 38 minutes of setup/teardown? [04:15] ugh, chromium keeps crashing in that tab [04:17] slangasek, ahh, that one is due to the prior job taking a long time (due to long tests). Jenkins can't complete build N+1 until build N is complete [04:17] um [04:18] well, that seems entirely unhelpful on jenkins' part :P [04:18] slangasek: I see what you mean by lack of timestamps. Thanks for confirming that [04:21] right, so http://s-jenkins.ubuntu-ci:8080/job/generic-deb-autopilot-runner-mako/3289/, the teardown is only 2 minutes, the autopilot run is 30 minutes, and the setup is 9 minutes [04:22] those numbers make more sense, at least [04:22] though confusingly, the first output at the top of the console log appears to be 10 minutes after the job was launched [04:36] bzoltan, I've deployed job updates to remove the coverage build for that branch [04:37] fginther: super! Should I expect that Jenkins will start to pick up the failed builds or should I rebuild the failing ones? [04:38] bzoltan, someone will need to restart the failed -ci builds. jenkins will pick up the jobs that failed -autolanding if the MPs are top-approved again [04:42] thanks fginther! an exploding bzoltan would be dangerous, and I live only 1 mile from him! [04:42] Mirv: :D === sergiusens_ is now known as sergiusens [05:03] tvoss: when you're around, https://code.launchpad.net/~thomas-voss/platform-api/fix-1348334/+merge/229816 + https://code.launchpad.net/~charlesk/indicator-location/lp-1348334-ualc-status/+merge/230004 need to be approved [05:03] for silo 014 [07:04] Mirv: would you assign a silo for the line 30? [07:08] sure [07:59] Mirv: I would need a reconfig for the silo2, i forget an MR to include [08:03] * ogra_ wonders why he cant reach launchpad [08:05] ogra_: works fine here [08:06] hrm [08:06] cihelp hey, I just discovered that http://s-jenkins.ubuntu-ci:8080/job/unity-phablet-qmluitests-utopic/? still doesn't load all our test results, it seems to only look at test*.xml files, where some of our files are actually *test.xml... [08:07] Saviq: i'll take a look at it in a bit. need to reboot [08:08] which is something of a problem that we need to resolve (basically redo our CMake around tests...), but in the mean time it'd be nice if it'd actually pick up all our results [08:08] psivaa, thanks [08:08] sil2100: I would need a reconfig for the silo2, i forget an MR to include. Would you be kind and please reconfigure the silo2? [08:08] psivaa, fginther said he updated the job like a week and a half ago, but that doesn't seem to have really happened [08:08] * ogra_ reboots his router [08:09] bzoltan, you can reconfigure yourself if you're not adding any new projects [08:09] ogra_, problems connecting to canonical's irc? [08:09] bzoltan, https://ci-train.ubuntu.com/job/landing-002-0-reconfigure/build?delay=0sec [08:09] Saviq: ahh, is that so? Is it a new feature? [08:10] bzoltan: was that a new project added or just a new merge? [08:10] bzoltan, no, been there since the beginning almost [08:10] sil2100: just a new MR [08:10] Right, a reconfigure from our side is only needed when you add a completely new project to the list, since this needs to be approved by the landing team (as it might conflict) :) [08:12] sil2100, good morning, can I have a silo for line 31 ? [08:12] sil2100, ogra_ - no meeting this morning i guess [08:12] brendand: what's wrong? [08:12] sil2100: I have never land mixed projects... it is either the UITK or the QtC plugin. I remember being rejected after I added a new MR [08:12] sil2100, can *you* access .canonical. or .ubuntu. sites? [08:13] brendand: launchpad works fine for me :) [08:13] brendand: the IRC as well [08:13] sil2100, launchpad hmm. that might be a different matter [08:13] bzoltan: you have to reconfigure then through the 'reconfigure' button [08:13] oSoMoN: sure [08:13] sil2100: I know that ... just wondering [08:13] sil2100, that's weird [08:13] funnily I also have network issues [08:14] bzoltan: I would, if I could reach the ci-train [08:14] Mirv: no need, I could do that [08:14] bzoltan: oh, ok [08:14] sil2100, thanks [08:14] ogra_, Mirv, brendand: so, if you guys would still have problems, I guess we'll skip the meeting [08:14] No need to have a hangout with just myself in it ;) [08:15] :) hangout seems to start alright, it's just a bit like half of the Internets would be broken [08:15] including LP + our VPN [08:15] but me and ogra_ shouldn't be sharing a router or connection :) [08:16] lol [08:16] sil2100, so odd you can access - there are even other people in poland who can't [08:16] sil2100, do you have some magic fingers? [08:17] hah! I've got some connection in teh goverments [08:17] Special red-line internet [08:17] ;) [08:17] a tracepath to ci.ubuntu.com ends after 8 hops with "no reply" messages for me [08:17] ogra_, I cannot even access canonical.com [08:18] right [08:18] ogra_, ends with "No Data Received" [08:18] ogra_, you should ask on #is in canonical internal... no wait [08:18] :( [08:19] I seem to reach the internal network ok [08:19] they are discussing the issue on #canonical-sysadmin on this IRC [08:20] hmmm, even my phone cannot reach the app store [08:20] indeed, why would that be different [08:20] tvoss, yeah i don't see updates [08:21] the meeting would be pretty pointless anyhow without access to any of our sites, even if hangouts would be working [08:21] heh ... topic in #canonical-sysadmin: "Known issues: Connectivity issues from some parts of Europe" :) [08:21] "some parts" [08:21] ogra_: yeah, just checked the same :) [08:22] "everywhere but sil21000's house" [08:22] i guess that should be "No connectivity issues from *some* parts of Europe" [08:23] pete-woods, around? [08:23] tvoss: yep [08:24] pete-woods, mind having a look over: https://code.launchpad.net/~thomas-voss/platform-api/fix-1348334/+merge/229816 [08:24] pete-woods, you haven't looked at it before, but I would like to the MP in and I addressed charles' comments [08:25] sil2100, so since we cant even see the dashboard (or changes) i guess we can skip [08:26] tvoss: I don't understand how you addressed charles' comments. there are no changes / comments following his. [08:26] pete-woods, yeah, sorry, just checked ... cannot push to launchpad [08:28] ogra_: ok, let's just discuss things on IRC then [08:28] ogra_: ...once you get your internet access back [08:29] lol, my internet is fine [08:29] or maybe sil2100 will quickly setup a server through which we all can route [08:30] tvoss: give me a shout when you manage to push your changes :) [08:32] landing call? [08:32] just me and psivaa on it, or are you all in another mirror called? [08:32] popey: where at [08:32] popey, we decided to skip [08:32] https://plus.google.com/hangouts/_/canonical.com/landing-meeting [08:32] oh [08:33] popey: I can't get on it [08:33] popey, due to the datacenter outage for most of us [08:33] ok [08:33] davmor2, indeed, your SSO account works via the datacenter :) [08:33] popey, don't tell me you can access stuff? [08:33] Ok, I won't. [08:33] lol [08:34] popey, O_o [08:34] (I can't) [08:35] thanks for skipping though, means I can eat my breakfast boiled egg in peace ☻ [08:35] finally a day for coding quietly in a corner \o/ [08:35] ☻ [08:36] ogra_, just don't try to push anything [08:36] oh, is this why i cant get #185 ? [08:36] indeed [08:36] popey, bingo [08:36] no new shiny for you til this is fixed [08:36] bummer [08:37] yeah, mean ... i'm still on 182 and see all these videos :( [08:38] heh, i saw a post from victor while I was in bed [08:38] thought .oO( I should update my phone ) [08:38] popey, damn ... stop mailing bug links around ... people click on them, yknow ? [08:38] hah [08:38] popey: already updated mine phew [08:40] * sil2100 cannot code in quiet as he has access to canonical servers :< [08:41] right back in a bit I'm going to finish my coffee while I wait for the data center to come back so I can find out if I'm meant to be doing something via email and bug reports [08:42] well, if it takes longer we should all consider visiting sil2100 [08:42] "wanna land something ? go to poland" [08:44] ogra_: that is not that far from here [08:44] :) [08:46] Mirv: I am good with the silo2 [08:48] Mirv: should we move jenkins to Poland too? http://s-jenkins.ubuntu-ci:8080/job/autopilot-testrunner-otto-utopic/2143/console [08:50] wouldnt that then be "s-kamerdyner.ubuntu-ci.pl:8080" ? [08:57] hahah [09:17] bzoltan: not approved https://code.launchpad.net/~zeller-benjamin/qtcreator-plugin-ubuntu/kitcreate/+merge/230286 [09:19] pete-woods, should be up to date [09:22] sil2100: hiya. How's silo 1 looking for landing? [09:22] (if the answer is "already done", whee :) ) [09:23] ;p [09:23] * Chipaca can't reach people.c.c [09:23] hm, maybe i can reach it from my french server [09:24] * jgdx can't connect to cilabs vpn === vrruiz_ is now known as rvr [09:25] so, I can get in via france [09:25] * Chipaca starts selling ssh access [09:25] for me everything is back up [09:25] oh! it's just come back, here [09:25] woo [09:26] * Chipaca closes his shop [09:26] Saviq: hey! === chrisccoulson_ is now known as chrisccoulson [09:26] Saviq: would you like to volunteer landing your recent unity8 landing for ubuntu-rtm? ;) [09:36] Mirv: approved [09:44] Mirv, both missing approves are in [09:48] sil2100, sure [09:49] sil2100, what's the process then? [09:49] sil2100, shall I branch lp:unity8/14.09? [09:50] (with whatever's in ubuntu-rtm at the moment)? [09:52] Saviq: yeah, I guess it would be good to create a branch there with 8.00+14.10.20140808-0ubuntu1 on top [09:52] bzoltan: tvoss: both done [09:52] sil2100, yup, doing [09:53] sil2100, do we have a convention yet on the series name? just 14.09? [09:53] Mirv: thanks [09:53] Saviq: from what I know it will be just 14.09, as the name is already solid in the archives [09:53] sil2100, ok, will have to do the same for the scopes which landed in the same silo [09:54] Saviq: ACK, thanks for volunteering! It might be a rough ride btw. [09:54] sil2100, /me likes a rough ride [09:54] Saviq: I'll use the preprod silo for now not to break the workflow of others [09:54] oops [09:54] did I just share too much? [09:54] uh oh! TOO MUCH INFO [09:54] ;) [09:55] hm, in the meantime I think I need to make a *real* debootstrap for 14.09 [09:55] davmor2: bug 1355700 [09:55] bug 1355700 in ubuntu-keyboard (Ubuntu) "Two spaces erroneously translates to fullstop-space every time" [Undecided,New] https://launchpad.net/bugs/1355700 [09:55] Saviq: 14.09 or rtm-14.09 for the branch name, I guess; you get to establish a convention :) [09:55] sil2100: You'll probably need ubuntu-keyring from ubuntu-rtm/14.09 [09:56] I've been trying to decide whether it's the right thing to do to add that to mainline ubuntu-keyring ... [09:56] Saviq: You can get the version to branch off from my recent mails to ubuntu-phone@, or just from https://launchpad.net/ubuntu-rtm/+source/ [09:57] cjwatson, yup [09:57] * Saviq goes for rtm-14.09 to be clear that this is re: ubuntu-rtm [09:58] Saviq: I guess you're setting up a productseries in LP as well with the new branch set as its focus? [09:58] Bit of a palaver but seems like the right thnig [09:58] *thing [09:58] cjwatson, yupps === psivaa_ changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | CI Train support: robru, trainguards | Vanguard (general help): psivaa_ | CI Train Status: #179 promoted | CI Train Dashboard: http://bit.ly/1mDv1FS | Known issues: - [10:01] popey: yeap confirmed, what was your one from last night [10:05] uh [10:06] bug 1355422 [10:06] bug 1355422 in unity8 (Ubuntu) "Can't dismiss notification bubbles" [Undecided,New] https://launchpad.net/bugs/1355422 [10:07] popey: ah that's the one thanks just tracked it down in my irssi emails === tvoss is now known as tvoss|test === tvoss|test is now known as tvoss [10:16] ^ it's in main, so a core-dev ack on the symbol adding at https://ci-train.ubuntu.com/job/landing-003-2-publish/66/artifact/packaging_changes_thumbnailer_1.1+14.10.20140811-0ubuntu1.diff would be useful for publishing [10:17] davmor2: hey! Could you perform promotion-wise dogfooding on our latest image? ;) #185 that is [10:19] sil2100: I can first I want to see if a bug I hit yesterday on my fresh flash is reproducible. I think the notification on mako for there is no storage is blocking the guide so you just get a black screen when you swipe the welcome screen away. [10:20] davmor2: oh, so that would mean that the issue that ToyKeeper also reported might be actually affecting someone, hm [10:21] cjwatson: is having the ubuntu-keyring from 14.09 safe for non-rtm purposes as well? [10:21] sil2100: yeah I was too busy trying to see about the no-sim issue which looks like it might be fixable [10:21] sil2100: Yes [10:21] Excellent [10:22] sil2100: I'll see if I can reproduce this and then carry on with a dogfood. the way to fix it is just to reboot [10:29] sil2100, hmm do we know what's gonna happen with translation updates? [10:29] Not sure yet... [10:30] sil2100, what I mean is that between https://launchpad.net/unity-scope-mediascanner/rtm-14.09 and the upcoming landing, there's a bunch of translation updates [10:30] sil2100, we probably need to register the series for translation, too [10:30] dpm: ^ any idea on translations wrt rtm? [10:30] cjwatson: hm, so, I tried creating a cowbuilder for 14.09 locally as a test run and it seems to be sad that 14.09 has no aptitude - do you know why we don't have it there? [10:30] https://launchpad.net/ubuntu-rtm/14.09/+source/aptitude <- as per here [10:31] sil2100: Nothing caused it to be included. Why not use sbuild? [10:31] We only copied a selection of packages. [10:31] Is aptitude going to be a problem for CI? [10:31] cjwatson: well, that's how CI Train does it, not sure if me changing this now is a good idea [10:31] Saviq: I just wonder how the touch langpacks would be built - separate packages + versions for rtm and non-rtm I guess, with at least selectively choosing 14.09 branches for rtm packages or something like that [10:32] cjwatson: I just know that I need a cowbuilder chroot now for building the source packages and without it in 14.09 I can't create it with --create [10:32] sil2100: Do you have a top-level list of the packages it attempts to use (no need to expand out dependencies, in fact I'd rather you didn't)? [10:32] Mirv, Saviq, sil2100, last time I spoke to wgrant about it, I understood that translations could be done on the derived distro, and that message sharing would work between Ubuntu and the derived RTM version [10:32] cjwatson: hmmm... let me try digging into that and getting that list [10:32] cjwatson, any idea on this? ^ [10:32] dpm, right, so we need to register the rtm-14.09 series for translations [10:32] dpm: I don't really understand LP translations [10:33] dpm: That sounds about right for what I remember the plan being, but I'm by no means authoritative [10:33] dpm: Translations aren't enabled for the RTM distro yet. [10:33] I need to adjust a script to do a partial copy first. [10:34] ok, thanks wgrant, cjwatson [10:34] I currently intend to just copy all the templates for the sources that we included in the copy. [10:34] wgrant, I'm registering the first rtm-14.09 series for unity8, unity-scope-{click,mediascanner,scopes}, please let me know what should I do to get translations going there [10:34] *serie? [10:34] series [10:34] sing. series, pl. series [10:35] I use serieses as the plural when it's ambiguous :/ [10:35] eg. variable names [10:35] wgrant: Hopefully for all the sources currently in ubuntu-rtm/14.09, as that'll change from the initial copy (e.g. the aptitude stuff above; possible gnutls26 removal after the next sync) [10:35] I rephrase when it's ambiguous :) [10:35] Right, all the sources that exist in the target. [10:36] (which also conveniently fixes the issue we have today where removed sources' templates persist forever) [10:36] Saviq, in principle you won't need to do much. The source package will have translations enabled for the rtm distro, and for the upstream series, we'll need to mark it for sharing translations with that source package and enable automatic imports/exports [10:36] Wiktionary does list a "serie" word in English but tags it as obsolete. It looks wrong to me as a native speaker. [10:36] Right. ubuntu-rtm's upstream links are copied from Ubuntu, but can diverge without a problem. [10:37] Sharing should just work AFAIK, but let me know if there are any oddities. [10:37] I hope to enable translations by Friday. [10:38] thanks wgrant [10:38] good to know [10:39] sil2100: okay so the guide does appear and on first swipe I get this http://davmor2.co.uk/~davmor2/screenshots-phone/device-2014-08-12-113801.png /me shakes his fist at Saviq I bet it's all his fault ;) [10:40] dpm, well, yeah, but for them to get synced to the series branch unless we set those up for translation (assuming that we want them to be synced to those branches) [10:40] pffff [10:40] davmor2, ouch, how do you get that? [10:40] Saviq: do a boot strap flash on mako, first swipe of the guide [10:41] davmor2, can you check ~/.cache/upstart/unity8-dash.log? [10:41] davmor2, unity8-dash is most probably not running at all [10:42] davmor2, anything in /var/crash? [10:43] Saviq: I'm wondering if it is the ciborium no storage announcement some how playing up the guide. I'm going to dig into the logs now though [10:43] davmor2: have you had an issue where Unity8 dash just froze? I am able to use the launcher but the unity8 dash and welcome screen just freezes and doesn't accept input at all [10:44] Saviq: unity8-dash.log looks to be filled with this UbuntuClientIntegration: connection to Mir server failed. Check that a Mir server is [10:44] running, and the correct socket is being used and is accessible. The shell may have [10:45] rejected the incoming connection, so check its log file [10:45] davmor2, ssooo, that looks like what people reported on flo [10:45] davmor2, that dash doesn't work on first boot [10:46] ogra_, was it you that reported ↑? do you have a bug#? [10:46] Saviq: _usr_bin_maliit-server.32011.crash [10:46] _usr_bin_unity8-dash.32011.crash [10:46] _usr_lib_arm-linux-gnueabihf_url-dispatcher_update-directory.32011.crash [10:46] _usr_lib_arm-linux-gnueabihf_url-dispatcher_update-directory.32011.upload [10:46] Saviq: I'll try and trigger a report on the unity8-dash crash [10:46] davmor2, initctl list-env -g | grep MIR please? [10:47] Saviq, davmor2, zyga reported one, i didnt get that because i dont wipe [10:47] bug 1355726 [10:47] bug 1355726 in ubuntu-system-settings (Ubuntu) " image 185 --wipe causes unity 8 not to show up on first boot (after welcome wizard) " [Undecided,Confirmed] https://launchpad.net/bugs/1355726 [10:47] Saviq: root@ubuntu-phablet:/var/crash# initctl list-env -g | grep MIR [10:47] root@ubuntu-phablet:/var/crash# [10:47] davmor2, not as root please [10:48] phablet@ubuntu-phablet:~$ initctl list-env -g | grep MIR [10:48] MIR_SERVER_NAME=session-0 [10:48] MIR_SERVER_PROMPT_FILE=1 [10:48] MIR_SOCKET=/run/user/32011/mir_socket [10:48] UNITY_MIR_SOCKET=/run/mir_socket [10:48] WIZARD_ORIG_MIR_SOCKET=/run/mir_socket [10:48] ah hmm [10:49] the WIZARD_ thing shouldn't be there I think [10:50] davmor2, initctl list-env unity8-dash | grep MIR please? [10:50] Saviq: is the Wizard the welcome screen to the guide. If it is the guide that is where it died. If it is the welcome wizard then no I guess it should of ended [10:50] davmor2, it's unrelated to the edge guide at all [10:50] davmor2, it's just that the dash didn't run after the wizard [10:51] Saviq: http://paste.ubuntu.com/8025664/ decided to stop spamming the channel :) [10:53] davmor2, and does unity8.log even mention an incoming connection? [10:55] Saviq: http://paste.ubuntu.com/8025693/ [10:55] davmor2, hmm ok, and then: [10:55] davmor2, initctl list-env unity8 | grep MIR please [10:56] davmor2, and does /run/user/32011/mir_socket exist at all? [10:57] Saviq: seems to http://paste.ubuntu.com/8025698/ [10:57] cjwatson: ok, so it seems to be a bit hard to get an exact list, but right now the builder fails while fetching this list of packages (it's the last step for the create operation): aptitude-common libboost-iostreams1.55.0 libcwidget3 libept1.4.12 libsigc++-2.0-0c2a libsqlite3-0 libxapian22 [10:57] sil2100: Can you show me a log file, or point at the code for what it's doing? [10:57] davmor2, oh [10:58] davmor2, no, it's all the same here where it works :| [10:58] * Saviq no get it [10:58] davmor2, `stop unity8-dash` and start unity8-dash`? [10:59] cjwatson: it's a binary, so I don't see what's happening exactly, but looking at the logs of utopic successful cowbuilder create operations, you can see what's installed when: https://ci-train.ubuntu.com/job/prepare-dist-debootstrap/11/console [10:59] cjwatson: you can see the currently affected step by moving to: "I: Obtaining the cached apt archive contents" in this log [10:59] sil2100: I was hoping for the code for the prepare-dist-debootstrap job, I think - I can look into cowbuilder independently [11:00] Saviq: http://paste.ubuntu.com/8025730/ I see blackness still no change [11:01] davmor2, ok, we'll have to dig more locally [11:01] davmor2, please reboot and see if you can spot a difference between the different commands when the dash ran fine [11:01] * sil2100 is not entirely informed of the steps cowbuilder and debootstrap perform [11:03] sil2100: I can figure that out. What I need to know is what the Jenkins stuff is running [11:04] The layer that's calling cowbuilder [11:04] cjwatson: sure, basically for ubuntu-rtm what is run is this: [11:04] DIST="14.09" HOME=~/citrain/chroot-tools/ sudo -E cowbuilder --create --debootstrapopts --variant=buildd --mirror=http://derived.archive.canonical.com/ubuntu-rtm/ [11:04] Jenkins is upsettingly opaque for this, but I presume it's in a bzr branch somewhere or something [11:04] Aha, thanks [11:05] I will chase down cowbuilder [11:05] It's just a job in jenkins.. we have no access to the instance below jenkins so we have a few jobs here and there deployed to do stuff for us ;) [11:05] We should eventually really make it use the current Launchpad chroots for the series/architecture in question and then use sbuild on top of that, launchpad-buildd style [11:06] But I agree, not right now [11:09] Saviq: after reboot I get the guide, I get unity8-dash and I see no Wizzard http://paste.ubuntu.com/8025782/ [11:09] davmor2, yup, expected [11:09] davmor2, can you remove ~/.config/ubuntu-system-settings/wizard-has-run === pete-woods is now known as pete-woods|away [11:10] davmor2, and reboot again, let's see if it's really first boot or wizard [11:10] sil2100: Where's the next step here? There's definitely more, because I see "pbuilder execute --buildplace /var/cache/pbuilder/build//cow.1439 --no-targz --internal-chrootexec chroot /var/cache/pbuilder/build//cow.1439 cow-shell /usr/bin/apt-get install eatmydata bzr-builddeb software-properties-common -yq --force-yes" being called and that doesn't come from cowbuilder --create AFAICS [11:11] Yes, after creating this we run one more command as well [11:11] Saviq, davmor2, i had issues in the past where the wizard didnt create ~/.pam_environment and the whole session went weird then [11:11] DIST=$DISTRIBUTION HOME=~/citrain/chroot-tools/ sudo -E cowbuilder --save --execute -- /usr/bin/apt-get install eatmydata bzr-builddeb software-properties-common -yq --force-yes [11:11] davmor2, check if that file exsists and has content [11:11] OK [11:11] * cjwatson runs derive-distribution to copy "build-essential aptitude cowdancer eatmydata bzr-builddeb software-properties-common" then [11:11] (but of course I didn't run this yet as the main --create fails) [11:11] davmor2: where's this bottom edge thing people talk about in the dash [11:11] (build-essential is probably already there, but) [11:12] popey: pissy little arrow at the bottom of the scope on proposed [11:12] popey, dash overview ;) [11:13] popey: hovering over ebay for me [11:13] i see no arrow [11:13] Saviq: so that worked for me this time [11:14] popey, and if you swipr from the bottom on the app scope ? [11:14] nothing happens [11:14] it scrolls [11:14] heh, interesting [11:14]