16:02:32 <scottda> #startmeeting cinder_testing
16:02:33 <openstack> Meeting started Tue Dec 20 16:02:32 2016 UTC and is due to finish in 60 minutes.  The chair is scottda. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:02:34 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:02:36 <openstack> The meeting name has been set to 'cinder_testing'
16:02:43 <eharney> hi
16:03:00 <scottda> As usual, no Agenda.....Just time to talk about anyone's test issues/patches/etc
16:03:17 <geguileo> hi
16:03:21 <erlon> scottda: I have some
16:03:35 <scottda> #link https://etherpad.openstack.org/p/Cinder-testing
16:03:43 <scottda> erlon: I figured you would :)
16:03:50 <erlon> scottda: :)
16:04:13 <erlon> #topic Multimatrix tests
16:04:27 <erlon> dont know if that works
16:04:57 <openstackgerrit> OpenStack Proposal Bot proposed openstack/python-cinderclient: Updated from global requirements  https://review.openstack.org/411072
16:05:02 <erlon> I have put some patches to allow testing migration (and other stuff in the future) with multiple backends
16:05:18 <scottda> Link?
16:05:20 <erlon> I need people to give a look o those
16:05:47 <erlon> scottda: https://review.openstack.org/381736
16:06:00 <erlon> scottda: cinder side
16:06:09 <erlon> scottda: https://review.openstack.org/381737
16:06:10 <scottda> #link https://review.openstack.org/381736
16:06:20 <erlon> scottda: project-config job
16:06:31 <scottda> #link Cinder test matrix https://review.openstack.org/381736
16:06:43 <erlon> scottda: another patch in devstack-nfs-pugin was merged todayl
16:07:01 <scottda> #link project-config job for test matrix of multi-backends https://review.openstack.org/381737
16:07:16 <openstackgerrit> Eric Harney proposed openstack/python-cinderclient: Use 'reset-state' for objects other than volume  https://review.openstack.org/413156
16:07:34 <erlon> for those who want to test it, you can do that by adding the project-config script into devstack-gate, in a dummy patch
16:08:28 <smcginnis> erlon: scottda would have to set the topic since he's the chair.
16:08:39 <scottda> erlon: cool, nice work
16:09:37 <erlon> scottda: thanks, I had a link for a run, but lost it now
16:10:23 <erlon> anyone have any question about?
16:12:02 <erlon> Next topic is about adding our in-tree jobs into Cinder-related jobs
16:12:26 <scottda> #topic Adding in-tree Cinder tempest to Cinder-related jobs
16:12:36 <erlon> https://review.openstack.org/#/q/project:openstack-infra/project-config+branch:master+topic:run-intree-tests
16:13:08 <erlon> some are already merged, but I need you guys to have a look and see if that is ok, and +1
16:13:33 <erlon> so people from infra feels more confident about approving
16:14:37 <scottda> erlon: Yeah, although they are asking why not just add to Tempest, and not push to Cinder in-tree first. I (and most of us) already know the answer to that....
16:15:14 <erlon> scottda: yep, it was good that he asked, I had that in my throat
16:15:16 <erlon> lol
16:15:18 <e0ne> hi
16:17:02 <scottda> erlon: Anything else?
16:17:24 <erlon> scottda: not on this topic
16:17:36 <scottda> erlon: Do you have another topic?
16:17:38 <erlon> scottda: next is just about the migration tests that are on tempest
16:17:52 * scottda is thinking that erlon should run this meeting....
16:17:58 <scottda> :0
16:18:10 <scottda> #topic tempest migration tests
16:18:25 <erlon> scottda: haha, you need to do it, so I don't talk alone
16:20:02 <erlon> so, smcginnis, if you can talk to the tempest ptl to push that
16:20:30 <erlon> im feeling bad for pinging then amost every day
16:20:39 <erlon> about hte same thing
16:21:04 <scottda> You know what we need....
16:21:17 <scottda> Someone on the Cinder team to be a Tempest/QA core member.
16:21:29 <scottda> *cough* erlon *cough*
16:21:50 <erlon> scottda: haha, yeap, I dont think they like me
16:22:25 <erlon> scottda: I prefer to use Cinder to host the tests
16:22:33 <scottda> Well, they are very short-handed. I think that's most of the issue with low velocity of patches.
16:22:34 <e0ne> erlon, scottda: I attend these meeting but not every. I should do it on a regular basis
16:23:03 <scottda> e0ne: Yes, I'm trying to attend every-other-one. The time zones are only friendly for you  or me every other week.
16:23:07 <erlon> e0ne: nice, good to have people involved here
16:23:36 <scottda> e0ne: This week is my turn 17:00 UTC
16:24:00 <e0ne> scottda: 1700UTC is aceptable for me too. added it to my calendar
16:24:20 <erlon> for me too
16:24:39 <erlon> scottda: are we alternating? I didnt' get this part in the last meeting
16:25:00 <scottda> e0ne: OK. I think the other time is 900 uTC
16:25:17 <scottda> erlon: No, not this meeting, the QA meeting ...Openstack QA
16:25:22 <e0ne> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting
16:25:30 <e0ne> "'Thursday alternating every week between 17:00 UTC and 09:00 UTC."
16:26:25 <erlon> scottda: e0ne: hmm, ok, Ill try to attend too
16:26:57 <scottda> erlon: Did you want to talk about Tempest Migration tests?
16:27:32 <erlon> scottda: hmm, I hit 2 bugs with the migration attached
16:27:40 <openstackgerrit> Merged openstack/python-cinderclient: Add Constraints support  https://review.openstack.org/406461
16:27:59 <erlon> scottda: didnt had time to dig into, 1 is that one I mentioned to you
16:28:31 <scottda> erlon: ahh..OK. Cinder migration tests. Got it. I haven't had time to dig in to that
16:29:30 <erlon> scottda: there are several bug filled in launchpad too, ill try to get to then when I have some time
16:30:13 <scottda> erlon: Yeah, that keeps getting starved out, but it is on my list.
16:30:42 <erlon> https://bugs.launchpad.net/cinder?field.searchtext=migration&search=Search&field.status%3Alist=NEW&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.status%3Alist=CONFIRMED&field.status%3Alist=TRIAGED&field.status%3Alist=INPROGRESS&field.status%3Alist=FIXCOMMITTED&field.assignee=&field.bug_reporter=&field.omit
16:30:42 <erlon> _dupes=on&field.has_patch=&field.has_no_package=
16:30:48 <erlon> ops
16:31:01 <erlon> http://bit.ly/2i6ekZg
16:31:10 <erlon> not all but several of this list
16:32:22 <erlon> Thats all for me
16:34:00 <scottda> geguileo: Are you busy? Want to talk a bit about HA/AA testing? I'm starting to look at drivers and how to test for comliance.
16:35:13 <geguileo> scottda: hi
16:35:35 <geguileo> scottda: what are you thoughts after looking at the drivers?
16:36:08 <scottda> geguileo: Hey. I haven't really started looking, it's on my agenda at my New Job to help review our drivers and see if we can certify that they are OK for AA/HA...
16:36:25 <scottda> geguileo: So, I figured that we'd look at the existence of local file locks.
16:36:45 <scottda> geguileo: Anything else that you know of that we should be concerned about? Other than extensive testing?
16:37:18 <geguileo> scottda: the problem is that they could be using local file locks if it's only a local mutual exclusion requirement
16:37:31 <geguileo> Since we can have multiple locking requirements
16:37:36 <scottda> right
16:37:38 <geguileo> In the cluster and in the node
16:37:51 <geguileo> So I don't see any way to check compliance beside actual testing
16:37:59 <geguileo> besides
16:38:05 <erlon> scottda: geguileo: I thougt we would consider all drivers as non-HA/AA and then driver vendors would need to change that
16:38:11 <scottda> geguileo: OK, I couldn't think of anything either. Just thought I'd see what you were thinking.
16:38:20 <geguileo> erlon:yes, that's already in the code
16:38:21 <scottda> erlon: That sounds correct to me.
16:38:28 <smcginnis> erlon: Yep, that's the last we discussed AFAIR.
16:38:56 <geguileo> But we should look for a way for vendors to do the testing
16:39:04 <geguileo> even if it's not in the gate at the moment
16:39:05 <smcginnis> So local locking is a definite no. Otherwise it's just a matter of testing and verifying there are no non-obvious issues, right?
16:39:16 <scottda> Yes, I'm asking as a vendor.
16:39:21 <geguileo> if we can automate it and all vendors use the same mechanism as the basis
16:39:29 <geguileo> it will probably help
16:39:36 <geguileo> even if it's self certification at the moment
16:39:58 <geguileo> smcginnis: local locking is allowed if you only need to lock in the node
16:40:12 <geguileo> smcginnis: for example because you have a library limitation or what not
16:40:18 <smcginnis> geguileo: Oh, good point.
16:40:32 <scottda> geguileo: OK, thanks. I'm just starting in on this. I'll share what I'm thinking when we talk about Cinder testing after the new year.
16:40:42 <geguileo> smcginnis: I still have to create a devref explaining all that stuff and how a/a works
16:41:34 <geguileo> scottda: my plan is to work on that in my Christmas holidays
16:41:46 <geguileo> test a couple of things I have in my head
16:41:56 <scottda> geguileo: :( Even Ebenezer Scrooge gave people the Holiday off ...
16:42:35 <geguileo> scottda: Oh, I have the Holiday off, but this looks like fun (as long as I don't look at how to integrate it with the gate)
16:43:04 <geguileo> scottda: And I have a long PTO from the 23rd until the 10th
16:43:06 <scottda> ha..yeah, gate testing will have to come later.
16:43:43 <geguileo> but let's sync up next year on this
16:43:57 <geguileo> because I think that's the next big thing in A/A once everything is merged
16:44:18 <scottda> geguileo: Cool. Will do. I'm finishing a re-run of the manual tests today. Then I can review (and I'll approve) some more patches.
16:44:36 <geguileo> scottda: Awesome!!!
16:44:38 <geguileo> thanks!!
16:45:03 <scottda> Anyone else have other Cinder Testing topics?
16:45:55 <smcginnis> *crickets*
16:46:19 <scottda> #endmeeting cinder_testing