17:19:24 <rohitk_1> #startmeeting
17:19:25 <openstack> Meeting started Thu Jul 19 17:19:24 2012 UTC.  The chair is rohitk_1. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:19:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:20:31 <dwalleck> I actually can't stay, but I wanted to prove that I was still alive and chugging along. I'll work on resubmitting my last instance validation branch and the Swift tests before the end of the weekend
17:21:13 <davidkranz> rohitk_1: I am working on the automated resource freeing thing to go in after your whitebox submission.
17:21:46 <dwalleck> Beyond that, I know there's been a lot of modifications around base test classes that don't match up with the Tempest code internally, so I'm going to have to figure out how I'm going to keep pushing tests without rewriting them
17:21:48 <rohitk_1> davidkranz: yeah, hope the gate jobs get fixed, they're preventing merges
17:22:04 <davidkranz> rohitk_1: Ok. I didn't know that :(
17:22:47 <rohitk_1> davidkranz: jaypipes is looking at it last what i know of
17:23:06 <davidkranz> OK, maybe that's why he is not here.
17:23:16 <rohitk_1> I wanted some suggestions on Cinder tests and how to approach them
17:23:43 <rohitk_1> i think we should have Cinder tests in parallel with it's newly forked development
17:23:48 <davidkranz> rohitk_1: Well, the claim was that cinder is not really an API change.
17:23:54 <rohitk_1> luckily api is the same
17:24:00 <rohitk_1> only endpoints change
17:25:00 <davidkranz> rohitk: I agree. The tests can be the same now and as cinder changes they can be adapted.
17:25:15 <rohitk> so I think there should just be a config option in Tempest to choose which Volume service needs to be tested
17:25:17 <davidkranz> rohitk: But this depends on whether volumes is going to be deleted from Folsom
17:25:21 <rohitk> n-vol/cinder endpoint
17:25:32 <rohitk> right, which is still under discussion
17:25:39 <davidkranz> If cinder is kept in Folsom we need to test both.
17:26:33 <rohitk> davidkranz: Would it make sense to have this utility to run existing volume tests against Cinder, in Tempest now?
17:27:15 <davidkranz> rohitk: Do you mean "ability to run ..."?
17:27:49 <rohitk> davidkranz: well, I just meant a way to configure tempest to run against a selected Volume service
17:28:07 <rohitk> depending on what's deployed in the devstack environment
17:28:26 <davidkranz> rohitk: I was thinking we would run both and just skip if one is not available.
17:28:47 <davidkranz> We couild make the volume client be a trivial subclass of cinder client.
17:29:03 <rohitk> davidkranz: existing tests should hold true for Cinder too, just the catalog type needs to change
17:29:20 <rohitk> tp
17:29:25 <rohitk> *to 'cinder'
17:29:35 <rohitk> davidkranz: agreed
17:30:26 <davidkranz> rohitk: OK
17:30:33 <davidkranz> Anything else for now?
17:31:01 <rohitk> davidkranz: no updates on the Fuzz test tool yet
17:31:12 <rohitk> so it seems
17:31:36 <davidkranz> rohitk: How is working on that?
17:31:47 <davidkranz> rohitk: Who I mean
17:32:05 <rohitk> davidkranz: pcrews??
17:32:18 <rohitk> davidkranz: not sure though
17:32:45 <davidkranz> rohitk: I guess Jay or Daryl know.
17:33:11 <rohitk> davidkranz: so until randgen is worked upon, are we accepting old style negative tests?
17:33:32 <davidkranz> rohitk: I don't think we should be spending time on them.
17:33:42 <rohitk> davidkranz: hmm
17:33:43 <Ravikumar_hp> we had put hold on negative test development
17:33:57 <davidkranz> Ravikumar_hp: Yes, we need to focus more on new Folsom features.
17:35:14 <rohitk> davidkranz: All right, thanks!
17:35:19 <davidkranz> The only real value I see is for targeted negative tests involving multi-tenant security issues.
17:35:35 <davidkranz> fuzztesting can't deal with that.
17:36:04 <davidkranz> But those are really positive security tests :)
17:36:51 <davidkranz> I have to go very soon. Is there anything else.
17:37:15 <rohitk_1> davidkranz: nope
17:37:23 <Ravikumar_hp> nope
17:37:41 <davidkranz> OK, then we can adjourn for now.
17:37:41 <rohitk_1> thanks all!
17:37:48 <rohitk_1> #endmeeting