[14:48] <hallyn> jamespage: hey, so I'd like to hook up an automated run of lp:~serge-hallyn/+junk/lxc-test
[14:49]  * jamespage looking
[14:50] <hallyn> it ain't purty, but catches bugs
[14:52] <jamespage> hallyn: so can it run in a KVM instance?
[14:52] <jamespage> or does it need to run on bare-metal?
[15:02] <hallyn> yup, it can run in kvm instance
[15:17] <jamespage> hallyn: sorry thinking about this - its not quite like anything else we already have automated
[15:18] <hallyn> jamespage: is there a url showing me how to make it more like the tests you have automated?
[15:19] <hallyn> jamespage: would the libvirt tests in lp:qa-regression-testing fit better?
[15:19] <jamespage> hallyn, not sure TBH
[15:22] <jamespage> hallyn: so we could make it a set of tests off the end of an install test
[15:22] <jamespage> like we do for lamp
[15:22] <jamespage> automated install of ISO with required packages and then execute your test.
[15:24] <jamespage> they run as root
[15:24] <jamespage> so that would be OK
[15:27] <hallyn> jamespage: that'd be great
[15:29] <jamespage> hallyn: OK so can you take a look at     bzr branch lp:~ubuntu-server-iso-testing-dev/+junk/server-tests-precise
[15:30] <jamespage> this is really a new test_case (see subdirectories)
[15:30] <jamespage> at the moment the tests that are run in each test case are python unittests
[15:31] <jamespage> which means we can grab the results and display in jenkins
[15:31] <jamespage> they must all be in a file called 'test' in the test_case directory
[15:31] <jamespage> you can handle some of the package install stuff using the preseed - see one of the other test cases for an example
[15:36] <jamespage> hallyn: do you think that you can refactor what you have into that framework?
[15:37] <jamespage> we can get it up and running for precise hoepfully
[15:38] <hallyn> absolutely - well, in time for precise will depend, but it can definately be refactored like that
[15:38] <hallyn> jamespage: thanks
[15:39] <jamespage> hallyn: I say for precise because I know the QA team are going to be working on a new test framework for delivery early next cycle
[15:43] <jibel> hallyn, if you can provide a test that is an executable and its output is junit xml then its good.
[15:43] <jibel> For precise we should be able to hack something into the existing tool like we did for software-center.
[15:47] <hallyn> (*$&%(*$&% xml *$&%*($&% )
[15:48] <hallyn> so to be clear, if i refactor it like this now, i'll have to do it again for p+1?
[15:48] <hallyn> if so, it may be a better use of my time to run the testsuite by hand and refactor for p+1
[15:49] <jibel> no, you wont because we'll use the same output format for p+1
[15:49] <hallyn> ah, ok
[15:50] <hallyn> thanks guys, made some notes, will looka t this after ff
[15:52] <rbasak> hallyn, did you still need me to run the lxc test on arm? I think you asked me to hold on the other day?
[15:59] <hallyn> rbasak: yeah that'd be great
[16:06] <jibel> jamespage, could you help verifying bug 708548 and bug 810068   in lucid ? I've no iscsi setup
[16:06] <ubot4> Launchpad bug 708548 in kickseed (Ubuntu Natty) (and 3 other projects) "kickstart not removing logical volume data (affects: 1) (heat: 11)" [Medium,Fix committed] https://launchpad.net/bugs/708548
[16:06] <ubot4> Launchpad bug 810068 in partman-iscsi (Ubuntu Natty) (and 7 other projects) "kickstart iscsi option broken (affects: 2) (heat: 16)" [High,Fix committed] https://launchpad.net/bugs/810068
[16:11] <jamespage> jibel: hmm - I might not have either - had a slight accident - let me see
[16:14] <jamespage> jibel: I think I can resurrect it but do we have any example kickstarts for testing this?
[16:18] <jibel> jamespage, I thought there was one attached to the report, but apparently not :/
[16:18] <jamespage> jibel: may make it a little difficult
[16:18] <jibel> indeed
[16:45] <jibel> jamespage, and could you test a "default" preseeded iscsi install ? do you have a preseed for that ?
[16:45] <jamespage> jibel: I don't have preseeds for any type of iscsi install
[16:46] <jibel> jamespage, you should fix that :)
[16:46] <jamespage> jibel: sorry - not trying to be difficult
[16:46] <jamespage> thats happening all by itself!
[16:47] <jamespage> jibel: I can easily setup a rig to let me do manual iscsi installs; still have scripts
[16:51] <balloons> weekly qa meeting in #ubuntu-meeting in 10 mins :-)
[17:02] <rbasak> hallyn: lxc-test.sh seems to be trying to download armhf oneiric which doesn't exist
[17:02] <rbasak> I had another weird error as well, so I'm reinstalling to see if it happens again
[17:05] <jamespage> jibel: when do these need to be verified by?
[17:05] <jamespage> I know i should know....
[17:06] <jibel> jamespage, should have been done already, but no later than tomorrow
[17:06] <jibel> if we want them in 10.04.4 images
[17:06] <jamespage> jibel: right - I see
[17:07] <jamespage> jibel: I can't do it right now but I can take a look first thing tomorrow? would that be OK
[17:08] <jibel> jamespage, sure, that would be.
[23:17] <bdmurray> if somebody could confirm bug 928447 that'd be neat
[23:17] <ubot4> Launchpad bug 928447 in ubiquity (Ubuntu) "possible to use a number for your computer's name (affects: 1) (heat: 6)" [Medium,New] https://launchpad.net/bugs/928447