=== maclin_ is now known as maclin
=== maclin_ is now known as maclin
[06:40] <pitti> Good morning
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
[07:50] <pitti> jibel: ah, so when I install lightdm-gtk-greeter, the greeter starts in the container with the dummy driver
[07:50] <pitti> jibel: unity-greeter crashes
[07:51]  * pitti tries to find something less intrusive, that greeter is in universe
[07:54] <pitti> hah
[08:05] <pitti> jibel: so I want to test this: http://paste.ubuntu.com/6918740/
[08:08] <pitti> oh, this already callsl dpkg-divert, I'll do that as well then (with _runInImage)
[08:11] <pitti> http://paste.ubuntu.com/6918768/ then
[08:18] <pitti> jibel: hm, it does the upgrade test on the original container, without an ephemeral overlay?
[08:19] <pitti> jibel: I had a bug in my code; how do I restore the original saucy container image now?
[08:19] <jibel> pitti, no, the overlay is created and mounted in the backend
[08:19] <pitti> sorry, I had assumed that would happen in overlays
[08:19] <pitti> jibel: ah, so I can just kill the job?
[08:19] <pitti> I lxc-attached to upgrader-ubuntu-saucy-trusty-desktop-amd64 and it showed me the in-progress upgrade
[08:20] <jibel> pitti, yes, did it not use an overlay?
[08:20] <pitti> jibel: it's currently running, so I can't tell yet
[08:20]  * pitti kills the job
[08:21] <pitti> jibel: ah, it did
[08:22] <jibel> pitti, phew :) it's in UpgradeTestBackendLXC, it creates a mount hook script called mount.sh and set it as mount hook
[08:22] <pitti> ah, I see
[08:22] <jibel> then the overlay is used when the container is started
[08:24] <pitti> yay, X is running now
[08:30] <jibel> Yay
[08:30] <jibel> pitti, thanks for fixing all this
[08:41] <pitti> jibel: c'est vert !
[08:42] <jibel> pitti, Cool
[08:42] <pitti> jibel: so I'd commit http://paste.ubuntu.com/6918860/ and roll it out, if it's ok for you
[08:43] <jibel> pitti, LGTM
[08:50] <pitti> jibel: done and rolled out
[08:50] <pitti> AutoUpgradeTester/UpgradeTestBackendQemu.py
[08:50] <pitti> +                                   "--addpkg", "linux-image-generic",
[08:50] <pitti> jibel: ^ is that something which we should commit to bzr? (that's a local diff on wazn)
[08:50] <pitti> I actually had expected that this is already installed by default
[08:52] <jibel> pitti, IIRC, it was not installed by vmbuilder
[08:52] <jibel> so, yes should be in bzr
[08:53] <pitti> jibel: oh, do we still use vmbuilder to create the VMs?
[08:53] <pitti> I'd actually say we should fail a test if linux-image-generic is missing
[08:53] <jibel> pitti, it is still there, hallyn is porting it to use uvtools instead of deboostrap
[08:56]  * pitti filed a bug report as a reminder
[08:57]  * pitti looks into that weird "old kernel" thing in the VMs
[08:58] <pitti> jibel: so perhaps vmbuilder still installs grub 1 into those?
[08:59] <jibel> pitti, yes, but I installs grub-pc and is supposed to migrate from grub-legacy
[09:00] <pitti> jibel: ah, so if the base image doesn't exist, it calls ubuntu-vm-builder to generate it; so I could do that locally to recreate that VM and see where things go haywire?
[09:02] <pitti> so grub 1 is "rc" (i. e. removed, but not purged)
[09:07] <pitti> jibel: so, calling "grub-install /dev/sda" works
[09:18] <pitti> jibel: just to understand this, ubuntu-vm-builder just gets you a bare-minimum VM; where does it install ubuntu-desktop?
[09:18]  * pitti doesn't see it in UpgradeTestBackendQemu.py
[09:28] <jibel> pitti, it is in the profile, DistUpgrade.cfg has an entry that point to a file with additional files to install during boostrap
[09:28] <jibel> bootstrap
[09:29] <pitti> ah, thanks
[09:29] <jibel> for example in upgrade-tester@aldebaran:~/auto-upgrade-tester/profiles/trusty/ubuntu-saucy-trusty-desktop-i386/DistUpgrade.cfg
[09:29] <jibel> there is
[09:29] <jibel> BasePkg = ubuntu-desktop
[09:29] <jibel> AdditionalPkgs = pkgs.cfg
[09:30] <jibel> BasePkg is a list of packages
[09:30] <jibel> AdditionalPkgs is yet another list of packages but in a external file
[09:34] <pitti> jibel: sheesh, ubuntu-vm-builder failed on my host, grep: /proc/cpuinfo: No such file or directory
[09:35] <pitti> forgets to mount /proc into the chroot or so
[09:36] <pitti> jibel: would it be ok if I modify the original VMs (run grub-install in those), and if that works, adjust vm-builder or UpgradeTestBackendQemu.py to do that?
[09:36] <jibel> pitti, it is okay, it is usually what i do
[10:36] <pitti> jibel: http://d-jenkins.ubuntu-ci:8080/job/upgrade-ubuntu-saucy-trusty-desktop-i386_vm/24/console \o/
[10:38] <jibel> pitti, world is becoming greener \o/
[10:38] <pitti> jibel: so I wonder, should I just add the grub-install command to UpgradeTestBackendQemu.py?
[10:39] <pitti> jibel: yeah, and I just sponsored/NEWed xorg-lts-transitional, which should hopefully make the 12.04.{1234} upgrades much better
[10:40] <jibel> pitti, I saw that, it's great. You can update the qemu backend with the grub-install command and we'll reprovision VMs
[10:41] <pitti> jibel: so for testing, I'd add that to the code, remove/rename the amd64 VM, and re-start the test; that should rebuild the VM, right?
[10:42] <jibel> pitti, right, remove the disk image and it will be bootstrapped again
[11:15] <davmor2> Morning all
[11:17] <slickymaster-job> morning davmor2
[11:26] <pitti> jibel: yay, upgrade tests also find actual bugs
[11:27] <pitti> aside from the backported X stack ones, bug 1270597
[11:27] <ubot5> bug 1270597 in postgresql-common (Ubuntu) "package postgresql-client-9.3 9.3.2-1ubuntu1 failed to upgrade: alternative pg_basebackup.1.gz can't be slave of psql.1.gz" [High,In progress] https://launchpad.net/bugs/1270597
=== _salem is now known as salem_
[12:56] <pitti> jibel: hm, why do we have ubuntu-desktop in a server upgrade test? (http://d-jenkins.ubuntu-ci:8080/view/Upgrade/job/upgrade-ubuntu-precise-trusty-server-lts-saucy-amd64/26/artifact/results/bootstrap.log)
[12:59] <jibel> pitti, interesting, it was not part of the original bootstrap.
[12:59] <jibel> pitti, and was not in yesterday's upgrade
[14:12] <pitti> jibel: hm, I don't remember logging into those containers, did I mess up anything?
[14:13] <jibel> pitti, there is no risk of messing up anything. I've been sidetracked by a HO, I'll have a look, maybe the original config is wrong
[14:13] <pitti> jibel: no worries, please don't interrupt anything
[14:14] <pitti> jibel: I'm still investigating the grub issue in VM, my initial fix didn't work
[14:24] <jibel> pitti, so from history.log the initial provisioning is wrong, but the profiles are correct
[14:24] <jibel> pitti, I'll recreate this container
[14:26] <jibel> and amd64 too
[14:29] <pitti> jibel: can you please poke me when these are done? my postgresql fix landed in trusty, I'd like to restart precise-trusty-server-tasks*
[14:29] <pitti> and the xorg fix is in as well, I'd like to run those too
[14:29] <jibel> you can restart *tasks* only p>t+lts-s had this problem
[14:29] <jibel> lts-s-server
[14:31] <jibel> pitti, I restart upgrade-ubuntu-precise-trusty-server-lts-saucy-(amd64|i386) you can restart tasks too
[14:31] <pitti> jibel: yes, done; thanks!
[14:32] <jibel> pitti, lxc being lighter than VMs maybe we could increase the number of executors to 4?
[14:32] <pitti> ah, http://d-jenkins.ubuntu-ci:8080/job/upgrade-ubuntu-saucy-trusty-desktop-amd64_vm/24/artifact/results/bootstrap.log grub-install was *still* grub 1
[14:33] <pitti> jibel: not sure, it's still rather heavy; I managed to kill wazn with a dist-upgrade in a container twice already
[14:33] <pitti> but you have a better gut feeling about what these machines can do
[14:34] <jibel> hm, the run queue is pretty low and there is not bottleneck on IOs
[14:36]  * pitti wonders why he doesn't see any of the print() statements of AutoUpgradeTester/UpgradeTestBackendQemu.py in bootstrap.log
[14:36] <pitti> oh, they are in the console log
[14:38] <pitti> grub VM fix, take 2 running
[14:39] <pitti> jibel: so ATM we have 1 slot on aldebaran apparently?
[14:39] <pitti> and two on wazn?
[14:41] <pitti> precise-trusty-server-lts-saucy-{i386,amd64} pass again
[14:42] <jibel> there are 2 slots on each hosts, so 1 free slot on aldebaran, 0 free on wazn and 2 free on alderamin
[14:44] <jibel> pitti, but these tests use a lot of disk so we cannot balance the load between all the hosts, it would duplicate the base images. So LXC are done on wazn, S->T in VMs on aldebaran and P->T in VMs on alderamin
[14:44] <pitti> *nod*
[14:45] <jibel> ideally we'd have 1 shared disk with all the base images and distribute the jobs on a pool of slaves
[14:45] <pitti> in this brave new future world of CI machinery we probably will
[14:46] <jibel> :)
[14:46] <pitti> jibel: btw, what is saucy-{i386,amd64}.qcow2 on aldebaran?
[14:47]  * pitti is a bit confused about all those images, bak/, etc.
[14:47] <pitti> or the 5 checkouts of auto-upgrade-testing
[14:47] <pitti> well, let's clean up next week
[14:49] <jibel> pitti, <release>-<arch>.qcow2 are the original images created by vmbuilder
[14:49] <jibel> pitti, we'll cleanup next week, remember this upgrader was not supposed to exist anymore and being replaced by utah :)
[14:56] <pitti> darn, upgrade-ubuntu-precise-trusty-server-tasks-* caught the previous postgresql-common still
[14:57] <pitti> will probably need some time for whichever mirror to catch up
[14:57] <pitti> err, http://d-jenkins.ubuntu-ci:8080/job/upgrade-ubuntu-precise-trusty-desktop-lts-saucy-i386/13/? -- quois ?
[14:57] <pitti> "quoi"
[14:58] <pitti> that's installing xserver-xorg-video-dummy
[15:17] <pitti> jibel: http://d-jenkins.ubuntu-ci:8080/job/upgrade-ubuntu-saucy-trusty-desktop-amd64_vm//25 ☺
[15:18]  * pitti cleans up the other VMs and restarts those
[15:20] <pitti> jibel: http://paste.ubuntu.com/6920445/
[15:20] <pitti> jibel: the first hunk was already there; I'll commit those as two commits to bzr, ok?
[15:21] <jibel> pitti, suonds good
[15:21] <jibel> ou
[15:24] <pitti> done, and rolled out
[15:33] <pitti> more green!
[15:56] <pitti> jibel: nice, http://d-jenkins.ubuntu-ci:8080/job/upgrade-ubuntu-precise-trusty-server-tasks-amd64/27/ survives the upgrade now, and detects a conffile prompt
[15:57] <pitti> although I can't actually see any prompt in the log
[15:58] <pitti> jibel: oh, I have an idea about http://d-jenkins.ubuntu-ci:8080/view/Upgrade/job/upgrade-ubuntu-precise-trusty-desktop-lts-saucy-i386/13/console
[15:58] <pitti> jibel: you can't actually install xserver-xorg-video-dummy in that precise, it would need to be *-lts-whatever
[15:59] <pitti> and apparently it's already installed anyway
[16:13] <jibel> I don't see where it got this bind9.dpkg-dist from, there is nothing in the logs
[16:18] <pitti> c'est une chose pour demain
[16:18] <pitti> I hope the two running jobs that I kicked off will succeed as well now
[16:27] <pitti> green!
[16:28]  * pitti starts the other precise.lts*->trusty desktop tests
[16:36] <pitti> jibel: as for bind, we kill the LXC overlay as soon as the test is finished, right? (for post-upgrade investigation)
[16:37] <jibel> pitti, right, and it's in tmpfs, I'll add an option to keep it and upgrade on a persistent directory
[16:38] <pitti> jibel: no worries, I'll try it manually tomorrow morning
[16:38] <jibel> pitti, okay
[16:56] <pitti> jibel: I pushed two new tests, FYI
[16:57] <pitti> this is fun!
[16:57] <pitti> ok, some red -> yellow, more on this front tomorrow
[16:57] <pitti> good night!
[17:02] <jibel> Good night pitti
=== bfiller is now known as bfiller_afk
[19:25] <balloons> evening dkessel
[19:25] <balloons> and moring to you Noskcaj .. how are you gents?
[19:26] <Noskcaj> morning balloons. Pretty good, still waiting for MOTU and my new pc(s).
[19:27] <dkessel> hey, good evening balloons :) i'm fine. how are you? how's the dog?
[19:28] <balloons> I'm doing well. The dog is up and moving about happily. He's all to happy to overdue it. My wife tells me I have to be careful; I like running, hiking, etc and the dog is always coming along :-)
[19:41] <dkessel> hehe :)
[19:42] <dkessel> mh. can't find the official instructions for ubuntu touch installation on the 2013 nexus 7.... i thought it has become the officially supported tablet now...
[19:45] <balloons> dkessel, it's "flo" right?
[19:46] <dkessel> balloons, "razor", I believe: http://en.wikipedia.org/wiki/Nexus_7_(2013_version)
[19:47] <balloons> regardless, the image is still wip.. however if you are feeling brave
[19:47] <dkessel> ...if somebody tells me that at least it won't render the nice new tablet broken.... ;)
[19:48] <dkessel> i chose the 32 gb variant and hope to get dual boot...
[19:48] <balloons> dkessel, https://lists.launchpad.net/ubuntu-phone/msg06352.html
[19:48] <balloons> rsalveti would know more
[19:49] <balloons> and, https://lists.launchpad.net/ubuntu-phone/msg06378.html
[19:49] <balloons> dkessel, ^^ that's probably better to use actually
[19:50] <balloons> I guess the tldr is if you wait a week or so, it'll be much easier
[19:50] <balloons> buonasera Letozaf_
[19:50] <Letozaf_> balloons, hello
[19:50] <Letozaf_> balloons, howzit?
[19:51] <balloons> Letozaf_, craving a cheese sandwich and some tea :-)
[19:51] <dkessel> hehe many thanks balloons :)
[19:51] <Letozaf_> balloons, lol I finished dinner about an hour ago
[19:52] <dkessel> balloons, I guess I can cope with stock android for some weeks at least ;)
[19:53] <dkessel> rsalveti, I will be glad to test any never installation instructions ;)
[19:54] <balloons> dkessel, no worries.. You'll be able to consume ubuntu touch soon enough
[19:56] <balloons> Letozaf_, do you have a favorite cheese?
[19:56]  * balloons is still stuck on cheese
[19:57] <Letozaf_> balloons, well not really I like a lot of types of cheese
[19:57] <Letozaf_> balloons, maybe I prefer Asiago it's an Italian cheese
[19:57] <balloons> mm yes, quite good..
[19:59] <Letozaf_> balloons, what kind do you prefer ?
[20:00] <balloons> last night I had some gouda with a bit of fruit..
[20:00] <Letozaf_> balloons, sounds delicious
[20:01] <Letozaf_> balloons, sometimes it's also nice to have  cheese with jam too
[20:01] <balloons> we must stop.. lol.. hunger
[20:01] <balloons> anyways, is AP working for you now Letozaf_ ?
[20:02] <dkessel> guys... i am getting hungry again :)
[20:03] <Letozaf_> balloons, haven't tired, but I will now, I was working on bug 1238902
[20:03] <ubot5> bug 1238902 in xorg-server (Ubuntu) "10de:0422 [Asus M2N] Xorg crashed with SIGABRT" [Medium,Triaged] https://launchpad.net/bugs/1238902
[20:05] <Letozaf_> balloons, looks like I'm getting good at compiling kernels :P (without configuration, just compiling :P)
[20:06] <balloons> Letozaf_, :-)(
[20:16] <rsalveti> dkessel: http://paste.ubuntu.com/6915695/
[20:16] <rsalveti> just make sure to boot the device at least once after doing oem unlock
[20:18] <dkessel> rsalveti, thanks. is setting up dual boot possible yet?
[20:18]  * Letozaf_ is updating autopilot
[20:18] <rsalveti> dkessel: yup, ondra knows how to do it
[20:19] <dkessel> ondra?
[20:20] <dkessel> rsalveti, ondra?
[20:21] <rsalveti> dkessel: he's in #ubuntu-touch
[20:21] <dkessel> ok thanks
[20:22] <Letozaf_> balloons, yeah! autopilot launch is working again :D
[20:24] <balloons> Letozaf_, excellent, so might we be able to finish reminders?
[20:24] <balloons> if I remember right, the issue was the app having changed
[20:25] <Letozaf_> balloons, yeah! think so, I am trying to remember
[20:26] <Letozaf_> balloons, maybe it's just better I update the one on my PC and run it to see what it's like now
[20:30] <Letozaf_> balloons, one thing I do not remember, now reminders app enters my Evernote account directly
[20:31] <balloons> Letozaf_, yes I think that's ok. We did work that would allow us to get around that.. my oauth-branches I believe use my account. We have a test account now.
[20:31] <balloons> anyways, first things first, let's make them work again ;-)
[20:32] <Letozaf_> balloons, but do I have to test using your test account or is it ok to use mine ?
[20:32] <balloons> Letozaf_, right.. test using your own
[20:32] <balloons> we'll do the work again to migrate to the special test account.. david set one up
[20:34] <Letozaf_> balloons, ok
[20:35] <Letozaf_> balloons, just one thing: reminders-app will enter the evernote account right away right? so I will have to fix the test that selects the account and then enters it
[20:38] <balloons> Letozaf_, umm.. I think if only one account is defined they might do that. But even if you have it all setup, you still need to select the account at the moment
=== salem_ is now known as _salem
[20:38] <balloons> not a bad idea to make that piece a little utility function you can remove later if needed
[20:39] <Letozaf_> balloons, I launched reminders-app with autopilot launch and it entered my account without me having to select it
[20:40] <balloons> Letozaf_, ahh then they already made the change :-)
[20:40] <Letozaf_> balloons, I will try to add a second evernote account to see what happens
[20:43] <Letozaf_> balloons, cool, yes if you have two accounts you have to select one, it you only got one it enters directly
[20:43] <Letozaf_> balloons, I will have to fix the test
[20:43] <Letozaf_> balloons, I will have to check if there are two accounts
[20:43] <balloons> Letozaf_, gotcha. Yea, I remember them wanting to change that, didn't realize it was done already. For the test then, yes assume we have 1 account already authenticated
[20:45] <Letozaf_> balloons, oh! ok, so I won't have to check if there are more than one, suppose we will always test with only one account
[20:45] <balloons> Letozaf_, yes.. We should develop a test for when there is a second account, but let's start simple :-)
[20:46] <Letozaf_> balloons, ok
[20:48]  * Letozaf_ need to reboot
[20:54] <Letozaf_> balloons, argh! I removed the second evernote account, but now reminders-app doesn't enter my account directly but requires me to select it so now I am not quite sure what is the right behavior
[20:54] <Letozaf_> balloons, maybe adding and removing the second account "broke" something
[20:54] <balloons> mmm.. tea.. the second cup is always better.. more steeping
[20:55] <balloons> Letozaf_, wild.. that might just be a bug actually
[20:55] <balloons> if you wipe the cache for the app, I assume it will fix itself
[20:55] <balloons> ?
[20:55] <Letozaf_> balloons, let me try
[20:55] <dkessel> balloons, nice to see you guys have got the google calendar with QA events online... that will make it easier for me to participate
[20:56] <balloons> dkessel, yes, It's been handy.. I hope you find it useful too
[20:59] <Letozaf_> balloons, weired, if you run reminders-app with autopilot launch it enters the one and only Evernote account direclty, If you launch autopiolot run -vv to run the tests, it wants you to select the only Evernote account
[20:59] <Letozaf_> balloons, I cancelled the ~/.local/share/com.reminders...    dir
[21:00] <balloons> Letozaf_, are you sure you are removing the directory? That sounds really odd
[21:00] <balloons> although honestly it should matter
[21:00] <balloons> one account should go straight in
[21:00] <balloons> I guess keep poking around till you figure out what's goin gon
[21:01] <Letozaf_> balloons, ok
[21:13] <Letozaf_> balloons, if I launch the reminders app installed on my PC fom touch coreapps ppa it doesn't enter directly the one and only account, is there a directory to remove so I can clear it's settings ? I do not think it's the ~/.local/share/com.reminders-app one as I have already tried removing it and nothing changed
[21:16] <balloons> Letozaf_, I suspect you are launching different versions of the codebase. What's in trunk should be what's in the ppa though
[21:17] <balloons> I wouldn't worry about removing the settings constantly.. It was just worth a try.. it should work, no matter what
[21:17] <Letozaf_> balloons, could be should I remove the ppa ?
[21:17] <balloons> Letozaf_, well you don't need to worry about the ppa version. Work using what's in the source
[21:37] <dkessel> good night. see you :)
[21:38]  * Letozaf_ is rebooting
=== psivaa is now known as psivaa-afk
=== bfiller_afk is now known as bfiller