=== _salem is now known as salem_ === salem_ is now known as _salem [05:27] thomi: hm, failure @ https://code.launchpad.net/~pitti/autopilot/print-tree/+merge/192134 [05:27] thomi: "W: Failed to fetch http://ppa.launchpad.net/autopilot/ppa/ubuntu/dists/trusty/main/binary-amd64/Packages 404 Not Found" [05:27] thomi: looks like a PPA bootstrapping issue? [05:29] pitti: yeah - I guess we're not building trusty packages yet? [05:29] thomi: I guess once the first MP lands, we would? [05:29] I'm officially EOD'd - are you able to poke that for me please? I guess it needs a config change in the CI team [05:29] chicken-egg :) [05:29] yeah, maybe' [05:29] thomi: no, that's "our" PPA [05:29] or maybe someone from the CI team can do something [05:29] thomi: I suppose I could just do a manual upload [05:29] oh, ok [05:29] might be best, just to get it started [05:30] thomi: a no-change upload of xpathselect as 1.4bzr38trusty0, sounds ok? [05:31] it's small, and the next commit will have a newer release [05:31] sounds good [05:31] it's only a PPA :) [05:33] up'ed [06:11] jibel: I can reproduce the dnsmasq "kills my eth0" problem locally with run-adt-test [06:12] jibel: for debugging this, it would be incredibly useful to have an easy way to log into the VM aside from ssh -- WDYT about providing a serial console with http://paste.ubuntu.com/6299102/ ? [06:12] $ nc -U /run/shm/adt/trusty-amd64*.img.serial [06:12] there I can log in, and have a shell [06:19] jibel: added documentation: http://paste.ubuntu.com/6299120/ [06:20] jibel: (fixed the "Right" typo locally) [06:56] Good morning [06:57] pitti, excellent. That's useful indeed. Otherwise you can connect with VNC but not always helpful depending on the state of the VM [07:03] jibel: oh, VNC? for an autopkgtest-like VM? [07:04] jibel: I thought that was a graphical thing only [07:04] jibel: ok to push? [07:05] pitti, it gives you a console when there is no graphical environment, but also a graphical environment when there is one. But that's probably overkill for autpkgtest because we never use graphical environment [07:05] pitti, ok to push [07:16] jibel: FYI, did a followup commit to explain how to do that with minicom (which looks much prettier0 [07:19] anyone confirm which jenkins instance is used by community? [07:22] iahmad, no jenkins instance is *used* by community, but results are publish to a public instance on jenkins.qa.u.c [07:23] jibel, I see, don't find if any of the community app tests are running on jenkins, unless I am looking at wrong place. [07:23] iahmad, does it answer your question or did you mean something else? [07:24] iahmad, which app phone apps or desktop apps? [07:24] jibel, phone apps [07:25] jibel, e.g. calendar, calculator, music etc [07:26] iahmad, results for these apps are published on the dashboard http://reports.qa.ubuntu.com/smokeng/trusty/touch/mako/5:20131024:20131024/4802/ for example [07:26] iahmad, let me check if jenkins jobs are published too [07:28] iahmad, https://jenkins.qa.ubuntu.com/view/Trusty/view/Smoke%20Testing/ I see calculator, calendar, ... on this page [07:29] iahmad, are they the jobs you are refering to? [07:29] jibel, right, exactly [07:31] iahmad: if you log in with your open id [07:31] you'll see which internal instance they refer to [07:32] and if you have the vpn set up, you can navigate to those from the dashboard [07:32] gema, I don't seem to have login account for it [07:32] I can login to the dashboard [07:32] iahmad: your same open id [07:32] that yo7u use for everything else [07:32] gema, ack [07:32] iahmad: I was referring to log in to the dashboard, we cannot log in on external jenkins, only retoaded can [08:03] pitti, for dh-python test, I think a restriction allow-stderr will fix it [08:03] jibel: how so? [08:03] jibel: this is complaining about some missing python2.6 stuff [08:04] and the test fails with code 2 [08:04] (I just ran it in a local trusty VM) [08:06] pitti, hm, http://10.98.0.1:8080/view/Trusty/view/AutoPkgTest/job/trusty-adt-dh-python/5/ fails with a code 4 and nosetests output on stderr === iahmad is now known as iahmad|afk [08:07] jibel: ah, that too; but it's the dsc0t-dh-python-stdout which is the "interesting" bit [08:07] jibel: but thanks for pointing out, that is missing as well [08:08] pitti, thanks for correcting me, I was looking at the console and saw all the tests passing :) [08:10] jibel: ah, the allow-stderr is already there [08:12] jibel: "fun" that it succeeds in an schroot runner, but fails in kvm [08:40] jibel: filed a debian bug for dh-python, I don't know how to fix it [08:41] jibel: until then we shouldn't let the new version near trusty and keep it in -proposed [08:41] as that smells like an actual bug [08:42] pitti, okay, thanks. [09:07] morning all [09:07] morning everyone [09:07] hey slickymaster [09:08] DanChapman: how are you? === vrruiz_ is now known as rvr [10:08] Morning all === _salem is now known as salem_ [13:05] fginther: did you sometimes observe timeouts in the otto tester? I got one in http://10.97.0.26:8080/job/autopilot-testrunner-otto-trusty/93/console [13:05] fginther: in the previous 4 or so builds I triggered today I got some actual ap logs; I'll just retry, but I wonder what went wrong here [13:06] fginther: it's the exact same code that worked before, I just fixed a race in ofono-phonesim-autostart [13:07] pitti, if something fails before it gets to the autopilot test is can just time out. [13:07] fginther: right, I was just wondering if that was an exceptionally rare case or something like that just happens from time to time [13:08] pitti, this used to hit us a lot when a package setup failed (which has been corrected) but there may still be a hole. [13:08] fginther: so, seems logind is working now, thanks to the upgrade; my initial debugging was just a red herring as I didn't actually see the called processes' stdout/err in the log; I adjusted the logging and now it seems fine [13:09] I was hoping that my ofono-phonesim-autostart fix was sufficient to make them all pass now [13:09] here's some finger crossing :) [13:09] pitti, oh that sucks, the jenkins timeout means it didn't gather any logs :-( [13:10] fginther: oh, you mean that they actuallly might have run at last, just got stuck in an infinite loop? [13:10] pitti, no, if you look at build 92, you'll see a bunch of artifacts, including some upstart logs which are useful for debugging why autopilot didn't start [13:19] pitti, https://bugs.launchpad.net/ubuntu-ci-services-itself/+bug/1244659 [13:19] Ubuntu bug 1244659 in Ubuntu CI Services "Upstream merger otto jobs don't collect logs when jenkins timesout" [Medium,Triaged] [13:20] fginther: ah, thanks [13:31] jibel: FTR, filed debian bug for trusty-adt-r-bioc-limma, -edger fails in the very same way [13:33] pitti, good, and the test is not very verbose [13:33] that's an understatement [13:33] jibel: it took me some 10 minutes to dig deep enough to get some useful log :) [13:34] good morning. [14:04] fginther: http://10.97.0.26:8080/job/autopilot-testrunner-otto-trusty/97/console finally got started, but it seems it's stuck again; so, apparently not a hitch [14:09] pitti, I've hopped on the machine to take a look [14:12] pitti, found an error in the gnome-session log, working on extracting it for you [14:14] fginther: oh, one which explains the "g_dbus_connection_call_internal: assertion 'object_path != NULL && g_variant_is_object_path (object_path)' failed"? [14:15] pitti, there was a dependency issue: http://paste.ubuntu.com/6300862/ [14:16] fginther: interesting, that's what I touched this morning [14:16] fginther: anything interesting in /var/log/upstart/ofono-phonesim.log ? [14:17] I tested the script in a VM, but it might fail for the same reason why it didn't come up in the previous tests (and thus triggered the "skip") [14:18] fginther: ah, but that's something which I could try on the otto test box through ssh, doing [14:18] pitt: [14:18] Waiting for modem to get powered... [14:18] mount: block device aufs is write-protected, mounting read-only [14:18] mount: cannot mount block device aufs read-only [14:18] WTH [14:18] pitti: ^ that's ofono-phonesim.log [14:18] pitti, this appeared on the console at about the right time: /var/log/syslog: Oct 25 13:55:48 trusty-amd64-20131023-0529 kernel: [247987.787023] type=1400 audit(1382709348.515:85): apparmor="DENIED" operation="mount" info="failed flags match" error=-13 parent=32720 profile="lxc-container-default" name="/" pid=32726 comm="mount" flags="rw, private" [14:19] ah, hah [14:19] jibel: ^ does that ring a bell? I think you recently investigated something in that direction? [14:20] fginther: oh, are you on 10.97.2.153 by any chance? [14:20] fginther: that has a running otto with the failed o-phonesim-autostart package [14:20] pitti, hm, nothing, where is this? [14:21] jibel: happens when trying to install ofono-phonesim-autostart into the otto CI test system [14:21] http://10.97.0.26:8080/job/autopilot-testrunner-otto-trusty/97/console [14:21] pitti, the same [14:21] fginther: ah, I had assumed that was a test box [14:21] fginther: so whenever I trigger something there in jenkins, I can actually debug it live on that box, nice [14:21] pitti, there a a few slaves that run the otto jobs [14:22] pitti, so check the job first to see where it was scheduled. And 10.97.2.153 was upgraded first to test the trusty transition [14:23] pitti, it will probably go back to a test/staging reserve next week [14:23] # mount --make-rprivate / [14:23] mount: block device aufs is write-protected, mounting read-only [14:23] mount: cannot mount block device aufs read-only [14:23] ah, so that's it [14:24] I did that mostly out of a habit, but we don't actually need it here [14:24] # start ofono-phonesim [14:24] ofono-phonesim start/running, process 10922 [14:24] \o/ [14:25] * pitti uploads to saucy [14:25] err, trusty [14:25] autofingers! autofingers! [14:25] fginther: I guess I can poweroff that container now, to free it for the next job? [14:25] pitti, ack [14:25] fginther: thanks for your help! [14:26] pitti, no problem [14:31] fginther: interesting, this time it got artifacts [14:31] fginther: oh, I suppose because I shut down the box manually before the timeout [14:31] pitti, right, it wasn't killed by the jenkins timeout hammer [14:32] now to wait for an hour for the fixed o-phonesim to land in trusty, and trying again :) === salem_ is now known as _salem === _salem is now known as salem_ [17:13] mzanetti: thanks for the authenticator app. [17:13] elopio: heh :) you're welcome [17:14] elopio: everything good or do you have suggestions for improvements? [17:15] mzanetti: no complaints. I have just tried it. Simple, pretty and it works :) [17:15] ok. thanks [18:26] evening [18:27] evening fetzed [18:28] evening fetzed [18:28] evening balloons [18:29] evening dan, [18:29] evening DanChapman .. ahh Friday Friday [18:29] anyone else installed 14.04 yet? launcher shows 13.10 haha [18:30] fetzed, yes, many things still say 13.10 :-) I'm on 14.04 [18:30] balloons, indeed and a quiet friday it seems to be. [18:31] yeah i got confused, it is a really early release :P [18:31] DanChapman, a bit.. a new automated test writer has emerged on the mailing list -- did you see? [18:31] I'm excited for some new folks.. [18:31] * DanChapman opens thunderbird [18:32] balloons, yeah be great to see some new people getting involved :-) [18:32] the idea for making roles was the plan. I can point people at that page for test writing and we can do a big push to ask people to get involved [18:33] we've plenty of time now to coach them along.. Asking for developers and writers is the next piece on the list [18:33] balloons, just so you know the gtk emulators MP should be done today, I got a bit held up yesterday. Just got some logging bits and the tests to finish off. https://code.launchpad.net/~dpniel/autopilot/autopilot_gtk if you want to browse [18:34] yeah i was going to ask balloons is there any tuts been done for coding autopilot tests yet, would like to learn and get invovled. [18:34] balloons, yes it seems have got the ML in full swing with folks chatting :-) I like the idea of roles, its easier for people to get involved in a more specific area they like [18:35] fetzed, have a look at this page: https://wiki.ubuntu.com/QATeam/Roles/TestWriter [18:35] nice one [18:35] for core apps / general autopilot knowledge there is this video: http://www.youtube.com/watch?v=qD_e_xqlBbg [18:36] Otherwise the official documentation plus the wiki links has nice tutorials [18:36] http://unity.ubuntu.com/autopilot/ [18:36] I would encourage you to dive in and start on something though.. Best way to learn is to try [18:36] we can definitely help [18:37] cheers ill have a blast in a few mins after eatin [18:40] fetzed, wonderful. DanChapman has you covered on any gtk autopilot stuff should you want to go that route.. [18:40] I can help as well -- what is peaking your interest? [18:43] ive always been interested in coding and normally work on php/html/java/sql ive done a tiny bit of python, id like o jump in an learn more coding langs plus if it helps in qa its even better [18:45] fetzed, yes, it's really cool stuff.. [18:46] I mean autopilot let's us write some neat tests. DanChapman for instance has been helping automate our image testing and has done a lot of great work in that area [18:48] fetzed, if helping out there interests you DanChapman put together a nice intro post: http://www.ubuntuhackers.com/ramping-up-automating-ubiquity/ [18:50] cheers ill have a look through these links [19:10] balloons, when is autopilot 1.4 planned to roll out? [19:12] DanChapman, it was out, but was switched back to 1.3 in trusty [19:12] DanChapman, work is happening right now to make the switch.. so *soon* [19:14] balloons, lovely cheers :-) === salem_ is now known as _salem [20:03] watched the youtube vid and had a look at the docs, going to have a practice at it and hopefully get some of the bugs sorted :P [20:04] fetzed, wonderful! [20:05] did it all make sense? :-p [20:05] yeah pretty much. Nothing looked to over complicated