16:00:41 #startmeeting Cinder 16:00:42 Meeting started Wed Dec 14 16:00:41 2016 UTC and is due to finish in 60 minutes. The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:43 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:46 The meeting name has been set to 'cinder' 16:00:47 Hey 16:00:50 hi 16:00:52 Hello. 16:00:55 Hi 16:01:03 dulek duncant eharney geguileo winston-d e0ne jungleboyj jgriffith thingee smcginnis hemna xyang1 tbarron scottda erlon rhedlind jbernard _alastor_ bluex patrickeast dongwenjuan JaniceLee cFouts Thelo vivekd adrianofr mtanino yuriy_n17 karlamrhein diablo_rojo jay.xu jgregor baumann rajinir wilson-l reduxio wanghao thrawn01 chris_morrell stevemar watanabe.isao,tommylikehu mdovgal ildikov wxy 16:01:09 viks ketonne 16:01:10 o/ 16:01:13 Hi 16:01:15 Hi! o/ 16:01:17 #topic Announcements 16:01:20 hi/ 16:01:25 morning 16:01:45 hi 16:01:50 It's a hemna! 16:01:51 o/ 16:01:56 hi 16:02:00 it is 16:02:01 hi 16:02:04 o/ 16:02:05 A hemna sighting in the wild! :) 16:02:08 hi 16:02:15 I love a wild hemna 16:02:20 #topic Announcements 16:02:28 Moving on quickly from that... 16:02:37 :-) 16:02:40 #link https://etherpad.openstack.org/p/cinder-spec-review-tracking Review focus 16:02:55 I'm cutting O-2 later today. 16:03:11 #link https://releases.openstack.org/ocata/schedule.html Schedule 16:03:20 smcginnis: what will be limmited drivers only? 16:03:23 This is our official driver and spec freeze. 16:03:26 hi 16:03:37 so the cutoff isn't friday ? 16:03:55 smcginnis: anything you want to get in before the cut? 16:04:00 smcginnis: ok, driver and spec 16:04:00 Hello :) 16:04:12 hemna: Nope, Thursday is the official OS one, the 14th was what we had published for Cinder. 16:04:42 ok, so I guess that begs the question about the XIV driver then 16:04:48 hemna: Yeah. 16:04:52 xyang1: There are a couple outstanding driver submissions (that all appear to want to be on the meeting agenda today) so we'll have to take a look at whether they're ready or not. 16:05:04 hemna: That's a refactor IMO, so not a new driver. 16:05:09 we haven't really been getting any reviews on it from others 16:05:37 smcginnis: Ok, so what is the cutoff date you are looking at for that? 16:05:41 we still need more unit tests, but the rest of the driver code can get some attention 16:05:49 hemna: I wanted to review it the other day but it had WIP on 16:05:56 jungleboyj: I guess that would fall under the O-3 one. 16:06:03 xyang1: Oh right, same. 16:06:08 hemna: Is that still WIP? 16:06:13 xyang1, we have unit tests coming from alon 16:06:19 OK, good. 16:06:35 so if you can ignore that for now, there still is a lot of other code to review 16:06:55 hemna: ok 16:07:01 Otherwise, there's still a lot of items on the list that we had identified as priorities that need attention, other than drivers. 16:07:18 Especially geguileo's long list of patches. :) 16:07:45 smcginnis: Should we do sign-ups for the HA patches, the way we do for drivers? 16:07:55 scottda: That may help. 16:07:56 scottda: +1 16:08:14 dulek has been doing a lot of testing on those I've seen. Would be good to get others as well. 16:08:15 smcginnis: the HA patches merge deadline is O-3, right? 16:08:21 xyang1: Right 16:08:45 smcginnis, scottda: the sign-ups do help with drivers 16:08:57 #topic Cinder testing meeting 16:09:03 scottda: hey 16:09:12 smcginnis, xyang1: we would like to discuss Veritas HyperScale status as well 16:09:20 Hey. So we used to have a Cinder testing meeting before this meeting, at 1500 UTC.... 16:09:42 But since we came of Daylight Savings, I've had a conflict. So has smcginnis and others seem to not be around... 16:09:43 ketonne: do you have successful CI reporting? 16:10:02 ketonne: Yeah, please don't put items in the announcements section. We do that topic later in the meeting. 16:10:13 So I was wondering if we should re-schedule that meeting? This is assuming we still want to talk about Cinder testing on a regular basis? 16:10:25 xyang1: CI is stable and reporting although we do have some failures which we are addressing 16:10:31 scottda: I do think it's good to discuss. 16:10:32 smcginnis: ok 16:10:37 scottda: +1 for a new time 16:10:55 scottda: +1 to re-schedule it 16:11:01 ketonne: noop tests are not stable. 16:11:10 after this meeting is a good time for me 16:11:20 scottda: what about 1600 UTC on Tuesdays? 16:11:32 Who is interested and in a UTC 0000 or UTC + X time zone? We want you to be able to come... 16:11:38 scottda: +1 16:11:41 e0ne: That should work for me. 16:11:43 e0ne: That works for me. 16:12:00 erlon: -1. it will be too late for me, especially in summer 16:12:21 e0ne: that would work for me, same time as Cinder meeting but Tuesday right? 16:12:27 e0ne: hmm 16:12:30 erlon: yes 16:12:33 smcginnis: We just verified today that CI is triggering for every checkin. This has happened a few hours ago. 16:12:59 e0ne: that works for me 16:13:05 Anyone have a Big problem with 1600 UTC Tuesdays? 16:13:38 eharney, xyang1, gouthamr, akerr, smcginnis, cFouts, e0ne, geguileo, dulek, flip214, ntpttr patrickeast, _alastor_, DuncanT, erlon, josemello,mdarnell, karthikp 16:13:45 YOu are all on the ping list for the meeting 16:13:47 scottda: +1 from me 16:13:49 i like 1600 UTC tues 16:13:50 it works for me 16:13:53 smcginnis: can you please specify the time of brunch creation? 16:13:53 Speak now, or we'll just do this... 16:13:56 scottda: 1600UTC works for me 16:14:04 Sounds good to me 16:14:04 Cibo: Not really 16:14:08 +1 16:14:11 scottda: 0000 UTC is late for me 16:14:17 scottda: I'm ok with it 16:14:35 scottda: sounds like a deal 16:14:36 OK, I'll change it on the meeting wiki and ping you all next Tuesday. Thanks everyone. 16:14:53 scottda: thanks 16:14:56 scottda: Thanks! 16:15:03 oh wait.. 16:15:03 scottda: Thanks! 16:15:03 smcginnis: I hope that I will be able to refactor code and publish + code review 16:15:19 Wasn't there something on the ML about not using Cinder room for meetings? 16:15:23 OR am I mistaken 16:15:34 I know they just added openstack-meeting-5 16:15:44 scottda: I think the decision was it's actually good to use the cinder room. 16:15:53 Cibo: dont forget the to fix the logs.gz downloads 16:15:54 Or at least that's the way it was leaning. 16:15:55 It's preferred that you use the meeting rooms, but it's not required. 16:15:58 smcginnis: OK, we'll keep it there. Thanks. Back to you... 16:16:12 scottda: Thanks! 16:16:38 #info Test meeting to be held in #openstack-cinder 1600 UTC on Tuesdays 16:16:48 #topic New driver submissions 16:16:55 I'm going to group these. 16:17:07 #link https://review.openstack.org/#/c/389622/ NEC driver 16:17:10 erlon: what logs.gz I do not archive any more http://openstack-ci-logs-open-e.s3-website-eu-west-1.amazonaws.com/log/86/386186/28 16:17:23 #link https://blueprints.launchpad.net/cinder/+spec/veritas-hyperscale-cinder-driver Veritas HyperScale driver 16:17:35 erlon: or I am missing something? 16:17:50 So from what I see, neither one of these is actually ready. :| 16:17:52 http://ci-watch.tintri.com/project?project=cinder&time=7+days 16:18:04 I would not consider either to have stable CI's at this point. 16:18:11 There are still issues being found in reviews. 16:18:37 And a lot of this should have been worked on weeks ago, to be honest. 16:19:03 On the plus side, it does seem both are being active and attentive to concerns. 16:19:24 So I'm not going to -2 them right now, but from where I stand it doesn't look good. 16:19:31 smcginnis: how many hours do we have until deadline? 16:19:41 Others are certainly encouraged to review and decide. 16:19:48 smcginnis: We are working on these actively for the Veritas HyperScale issues reported. The last one by xyang1 was addressed just before this meeting 16:19:58 e0ne: I was hoping to submit the o-2 request after lunch. 16:20:12 But I can wait until later. 16:20:20 smcginnis: what timezone are you in? 16:20:30 Central US 16:20:35 thanks 16:20:38 Currently 10:20 AM 16:20:51 smcginnis: thank you 16:21:07 My biggest concern for both right now is the CI. 16:21:25 They are being stated as being stable, but that's not what I'm seeing. 16:21:40 I'm sure DuncanT would have an opinion on that too. :) 16:22:03 any thoughts on the Veritas Access driver? 16:22:18 https://review.openstack.org/#/c/409914/ 16:22:29 smcginnis: there is 1 day and 4 - 5 hours left... 16:22:40 sivn: Again, no CI. Driver submitted just days before the deadline. 16:22:50 Cibo: No, there is more like 4-5 hours left. 16:23:04 smcginnis: I see 16:23:45 stable CI reporting for Veritas HyperScale should be seen in a few hours. We fixed the last known issues last night India time 16:23:55 I'll step back and let others comment. This isn't just my decision, just stating what I see based on what we've published as our requirements and our deadlines. 16:24:14 so the original patch was submitted on Nov 27 16:24:39 hemna: Which one? 16:24:51 https://review.openstack.org/#/c/382097 16:25:15 That's the HyperScale one. The Access one is different. 16:25:19 I haven't looked at it at all because it hasn't had CI 16:25:33 hemna: Oct 4 was the first submission 16:25:38 hemna: So, that is ok ... 16:25:38 ok sorry, wasn't sure which one we were talking about 16:25:42 oh 16:25:49 Veritas access was just submitted 16:25:50 :( 16:25:51 Looks like they have been responsive to reviews. 16:26:03 True. That is good. 16:26:06 I'd say no on the Access driver 16:26:25 the deadline is for merging the driver not submitting afaik. 16:26:30 To give some context on Access driver, this was checked-in to kilo release but was removed due to lack of CI, we are resubmitting it 16:26:32 hemna: +1 16:26:39 hemna: oh, I see what you meant. the access one was too late 16:26:46 yah 16:27:42 the unit tests are lacking as well 16:28:22 well I guess the driver is a bit light, because it's mostly an nfs driver 16:28:25 I have uploaded one set of test cases? 16:28:30 sivn: does your driver do any call to the storage? 16:28:32 anyway, I think it's submitted too late IMHO 16:28:57 hemna: I think we should discuss about the drivers on the review track etherpad first 16:29:02 sivn: it seems only a layer over the NFS driver and that does not do anything optimized 16:29:08 for snapshot of the volume, it runs a specific command over NFS 16:29:10 hemna: +2 16:29:30 otherwise uses most of the NFS driver capabilities 16:29:37 sivn: ok, ill let a comment there and yu can follow up 16:29:53 thanks 16:30:13 we are working on CI too, that is in the last stages 16:30:31 the clone volume doesn't use a new size either. 16:30:34 we have got experience now, with the Veritas Hyperscale CI :) 16:32:00 OK, please all cores, take a look and chime in on the driver reviews in the tracking etherpad. 16:32:56 smcginnis: so we have two new drivers that might make the cut? NEC and Veritas HyperScale? 16:33:24 xyang1: That's my take, but a little iffy. 16:33:40 Getting back to HyperScale, last comment has been addressed and new test cases have been added as per xyang1's comments 16:33:43 smcginnis: ok 16:33:47 I will take a look at both soon. I do think they are close and they have been responsive to review feedback. 16:34:03 CI is really the biggest issue all around here. 16:34:13 smcginnis: I haven't seen CI yet, so that is the concern 16:34:27 yah CI is supposed to be reporting if you want a driver to merge 16:34:34 and passing :) 16:34:34 4+ hours isn't much time for CI. 16:34:51 CI's in general are pretty miserable as it is: http://ci-watch.tintri.com/project?project=cinder&time=7+days 16:34:59 hemna: haha yeah passing too. 16:35:09 diablo_rojo_phon: yeap, it talks aroung 1+ hour just to run a job and detect a failure 16:35:42 Ouch. 16:36:20 #topic Open Discussion 16:36:30 we are on the CI issues. you should see stable CI from HyperScale by today. Failures are being looked into on priority 16:36:30 Anything else to discuss? 16:36:50 smcginnis: when are we going to start working on topics list for PTG? 16:36:54 JovianDSS driver 16:36:57 eharney, thanks for the review man. much appreciated. 16:37:00 ketonne: That's good. But again, we should see stable CI _prior_ to the deadline, not on the deadline. 16:37:13 e0ne, after xmas ? 16:37:14 e0ne: Oh, good point. We should probably get an etherpad started for planning that. 16:37:31 so I am working on fixing existing issues... anything else? 16:37:33 Which reminds me - Summit session proposals will be opening up very soon. 16:37:54 Good time to start thinking about what you might want to submit for talks at the summit. 16:38:14 That will be coming up fast. 16:38:15 hemna: I'll have Orthodox Christmas after it:) 16:38:43 diablo_rojo_phon: How we on Cinder registration for the PTG? 16:38:54 :) 16:39:13 I didn't register yet. I'm going to do it next 2 weeks 16:39:21 e0ne: Great! 16:39:34 I'm still trying to get approval.... 16:40:29 I am registered. May be sending myself, but I will be there. 16:40:34 #link http://www.openstack.org/ptg PTG registration 16:40:48 :/ 16:41:00 smcginnis: still on the lower end. 16:41:03 There is the travel assistance program for this as well. 16:41:17 e0ne: I didn't see you apply for the tsp 16:41:36 diablo_rojo_phon: I've got approval from my company this time :) 16:41:38 OK, I guess that's all I've got. Anything else? 16:41:51 diablo_rojo_phon: so I don't need TSP 16:42:06 Woo-hoo! 16:42:27 OK, thanks everyone. Go forth and review. :) 16:42:31 smcginnis: should we shoot for o-3 for HyperScale then? 16:42:46 ketonne: What do you mean? 16:43:19 smcginnis: we shall, our beloved ptl. 16:43:31 do we have the window to get into ocata in O-3 for hyperscale? 16:43:40 ketonne, the deadline for new drivers is O-2 16:43:41 ketonne: I'm afraid, that you have to do it for P-2 :( 16:43:55 you've missed the deadline. 16:44:02 ketonne: No, the new driver deadline for Ocata is today. Otherwise you're looking at Pike. 16:44:46 OK, thanks everyone. 16:45:01 #endmeeting