17:03:25 <davidkranz> #startmeeting qa
17:03:26 <openstack> Meeting started Thu Sep 27 17:03:25 2012 UTC.  The chair is davidkranz. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:03:27 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:03:28 <openstack> The meeting name has been set to 'qa'
17:04:18 <davidkranz> mnewby: You there?
17:04:27 <mnewby> davidkranz: here
17:05:03 <davidkranz> OK, I don't see Daryl or Jose so we are probably not going to get a swift update.
17:05:35 <davidkranz> mnewby: How is the quantum stuff going? I replied to some one asking about quantum tempest testing on the openstack list.
17:05:53 <davidkranz> Do you know him or did he just arrive?
17:06:28 <mnewby> davidkranz: I'm afraid the effort stalled while I worked on getting quantum running reliably with devstack.
17:06:56 <davidkranz> mnewby: OK, that certainly has to come first :)
17:07:25 <mnewby> davidkranz: devstack/kvm+quantum is more or less working.  working on devstack/xen+quantum now.
17:07:48 <mnewby> davidkranz: once I finish that I'll get back to tempest tests targeting a minimal devstack config, as a starting point for more involved testing.
17:08:15 <davidkranz> mnewby: Sounds good. At that point other people could add tests as well.
17:08:27 <davidkranz> mnewby: Perhaps there should be a blueprint for this.
17:08:52 <davidkranz> mnewby: I saw one for quantum testing in the tempest area but it seemed much older.
17:09:08 <mnewby> davidkranz: I'm open to that.
17:09:51 <davidkranz> I think blueprints make it easier for other people to jump in without stepping all over each other.
17:10:20 <torandu> +1
17:10:52 <davidkranz> torandu: I was out yesterday but saw there was an IRC discussion of the volumes flaky failures.
17:11:00 <davidkranz> Is anything more known about that?
17:11:24 <davidkranz> torandu: I think Jay was involved in that discussion.
17:11:32 <torandu> no, I don't have any info
17:12:05 <torandu> I met w/ jay this am to discuss devstack & expectations re testing
17:12:11 <torandu> I've had some issues w/ my env
17:12:26 <torandu> I'm going to collect output and post for review
17:12:38 <davidkranz> torandu: OK
17:13:01 <torandu> I'm still green and want to make sure my env is correct
17:13:20 <torandu> Jay tells me that exercise.sh should all pass. that hasn't been the case for me
17:14:03 <davidkranz> torandu: Gool luck.
17:14:28 <davidkranz> Are any of the IBM guys here?
17:15:02 <mtreinish> davidkranz: I'm here
17:15:49 <davidkranz> mtreinish: I think the existing XML changes have all been approved and are just being held up by needing to be rebased.
17:16:05 <davidkranz> mtreinish: Are there any more coming?
17:16:48 <mtreinish> davidkranz: I don't think there are any more right now, we've mostly been doing the samples api work in nova lately
17:17:07 <davidkranz> mtreinish: OK, sounds good.
17:17:47 <davidkranz> Does any one have any other suggestions for summit sessions?
17:19:20 <torandu> davidkranz: what's been proposed so far?
17:19:45 <jaypipes> hey guys, here now. sorry for being late
17:20:19 <davidkranz> torandu: See http://summit.openstack.org/, sort by topic and look at QA
17:20:28 <davidkranz> jaypipes: Hi Jay.
17:21:04 <jaypipes> davidkranz: ty for taking up the summit stuff. really appreciated.
17:21:38 <davidkranz> jaypipes: No problem. I have been swamped with other stuff and this is a good excuse to get back to some tempest stuff.
17:22:17 <davidkranz> jaypipes: Monty proposed http://summit.openstack.org/cfp/details/69 for the QA session.
17:22:34 <jaypipes> lemme take alook
17:22:37 <davidkranz> jaypipes: He said that stuff supports parallel execution....
17:23:00 <davidkranz> jaypipes: Perhaps we could kill two birds with one stone.
17:23:49 <davidkranz> davidkranz: Getting help from the ci team on this would be huge.
17:24:26 <jaypipes> davidkranz: yeah, I'd support that. I've looked at testtools and found it nice, but not sure it will work great with the kind of module and package-level fixtures we use in Tempest. though... to be fair, nosetests doesn't either :(
17:25:07 <davidkranz> jaypipes: Presumably there is another way to do that if necessary.
17:25:58 <jaypipes> davidkranz: perhaps. we can also look at doing something home-grown, like the CloudScaling tarkin project: https://github.com/cloudscaling/tarkin
17:27:04 <jaypipes> davidkranz: as for the volumes stuff, jgriffith was investigating and will let us know his findings
17:27:19 <davidkranz> jaypipes: Good. That seems like a really nasty issue.
17:27:37 <jaypipes> davidkranz: not sure yet :( I haven't seen any failures myself
17:28:13 <davidkranz> jaypipes: I can't get one locally either but they have happened a lot in the gate.
17:29:04 <jaypipes> davidkranz: yeah, I was asking jgriffith for a link to failures... couldn't find any. but at the rate in which commits are flying through, not surprising. if you have a link, please share
17:29:27 <davidkranz> jaypipes: With regard to the parallel stuff, I thought there would be a big advantage to doing what the ci group does about this.
17:29:44 <jaypipes> davidkranz: agreed.
17:29:44 <davidkranz> jaypipes: I filed a bug about the volume thing and put in a link. I'll chase it down in a sec.
17:30:28 <davidkranz> jaypipes: Here it is https://bugs.launchpad.net/tempest/+bug/1056213
17:30:31 <uvirtbot> Launchpad bug 1056213 in tempest "Flakey build failures in some volume tests" [Undecided,New]
17:30:43 <jaypipes> cheers
17:31:00 <jaypipes> I'll try to take a look at it (though it's a bit tough to find time right now :(
17:31:26 <jaypipes> how are we doing as a team on code reviews in the tempest code review queue?
17:31:36 <davidkranz> One other thing we have to think about is whether to continue work on stable/folsom for now (need to create it anyway) or on master (grizzly)?
17:32:22 <davidkranz> davidkranz: I think I would favot stable/folsom for a bit.
17:32:36 <jaypipes> davidkranz: well, I think it's good to work on stable/folsom until the release at a minimum...
17:32:46 <jaypipes> davidkranz: talking to yourself again? ;)
17:33:07 <davidkranz> jaypipes: Sticky fingers :)
17:33:17 <davidkranz> jaypipes: Wasn't Folsom just released?
17:33:25 <jaypipes> davidkranz: RC I think?
17:34:12 <davidkranz> jaypipes: Thierry sent a message. I think it was the real thing
17:34:50 <jaypipes> oh, ok. then definitely, we need to make the stable branch
17:35:10 <jaypipes> davidkranz: I will work with jeblair on that.
17:36:16 <davidkranz> jaypipes: OK.
17:36:56 <davidkranz> Anything else, any one?
17:38:14 <jaypipes> not from me. I'll look over the design summit proposals today and get someinput to you
17:38:45 <davidkranz> jaypipes: OK, thanks. Some of them are obviously not appropriate for the QA session.
17:39:15 <torandu> when is the list finalized?
17:39:48 <davidkranz> torandu: It is supposed to be finalized by end of next week but I would like to have all the proposals in by Wednesday.
17:40:13 <torandu> ok. I'll catch you off-meeting for more detail. thx
17:40:16 <davidkranz> torandu: I am trying to grab two more slots from swift or horizon.
17:40:44 <davidkranz> OK. In that case I will end the meeting.
17:40:52 <notmyname> I should know more by the end of monday wrt swift slots
17:41:10 <davidkranz> notmyname: Ah. I was just going to send you an email.
17:41:38 <davidkranz> notmyname: While you're here, do you know what happened to the tempest swift tests?
17:41:54 <jaypipes> davidkranz: notmyname is no longer at RAX
17:42:06 <notmyname> davidkranz: last I knew they expired (but that was several weeks ago)
17:42:12 <notmyname> expired in gerrit
17:42:17 <jaypipes> yeah.
17:42:18 <davidkranz> jaypipes: I know, but he is still PTL...
17:42:28 <jaypipes> davidkranz: :) understood
17:42:34 <davidkranz> davidkranz: OK, I'll ask Daryl.
17:42:45 <davidkranz> Talking to myself again dammit!
17:42:55 <notmyname> I've been waiting for them to get in so that I'm not dependent on RAX QA schedules for swift releases
17:43:10 <jaypipes> notmyname: ++\
17:43:22 <jaypipes> they've been expired and restored about 10 times now :(
17:43:30 <davidkranz> notmyname: That relates to the topic I just proposed for the summit about "bringing QA into the open".
17:43:40 <jaypipes> the things we needed cleaned up really weren't that much, either...
17:44:00 <davidkranz> notmyname: RAX must have more tests than were represented in JOse's submission.
17:44:31 <notmyname> davidkranz: RAX's tests also cover end-to-end tests for the value add (CDN) and product (billing) stuff
17:44:45 <torandu> davidkranz: "bringing QA into the open" has been on my mind. +1 on the topic for discussion at the summit
17:45:46 <davidkranz> notmyname: We really want tempest to be an "acceptance test" for openstack which is I guess what you want too.
17:46:39 <notmyname> davidkranz: my main concern is that releases have enough QA to be prod-ready. getting the tests into tempest is one good way to do that.
17:47:27 <davidkranz> notmyname: Yes, that is the subject to be discussed at the summit: how to make that happen.
17:47:48 <notmyname> I also want to run the swift tests here at swiftstack (for our product releases and for swift releases)
17:48:21 <notmyname> again, tempest is one way to do that and probably one of the easiest ways to do it
17:48:34 <davidkranz> notmyname: Of course. I just thought it would be good if "the swift tests" you mean were in a pubolci repository.
17:48:54 <davidkranz> Grr. "public repository"
17:49:11 <davidkranz> I think we are all in agreement about this.
17:50:09 <notmyname> yes
17:50:44 <davidkranz> I have to run to another meeting so I am going to end this for now. I will update my summit proposal a little bit based on this discussion.
17:51:07 <jaypipes> davidkranz: cool, thx man
17:51:18 <davidkranz> Talk to you all soon.
17:51:25 <davidkranz> #endmeeting