17:00:09 <alaski> #startmeeting nova_cells 17:00:10 <openstack> Meeting started Wed Jul 29 17:00:09 2015 UTC and is due to finish in 60 minutes. The chair is alaski. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:11 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:14 <openstack> The meeting name has been set to 'nova_cells' 17:00:49 <alaski> anyone here today? 17:00:55 <vineetmenon> o/ 17:00:58 <melwitt> o/ 17:01:00 <belmoreira> o/ 17:01:12 <bauzas> _o 17:01:16 <bauzas> \o even 17:01:43 <alaski> great 17:01:53 <alaski> #topic Tempest testing 17:02:02 <alaski> melwitt: anything to report? 17:02:18 <melwitt> alaski: no, don't think so 17:02:35 <bauzas> we had some problems with the devstack install last night 17:02:41 <bauzas> (not speaking of tempest) 17:02:52 <bauzas> uh, last night EU time 17:03:06 <alaski> what sort of problems? 17:03:16 <bauzas> alaski: sec, giving the URL 17:03:26 <bauzas> alaski: nothing related to cells, rather timeouts AFAICS 17:03:43 <bauzas> anyway, that's the only issue I spotted this week 17:04:24 <bauzas> http://logstash.openstack.org/#eyJzZWFyY2giOiJtZXNzYWdlOlwiZXhpdCBjb2RlOiAxMzdcIiBBTkQgdGFnczpcImNvbnNvbGVcIiBBTkQgYnVpbGRfbmFtZTpcImdhdGUtZGV2c3RhY2stZHN2bS1jZWxsc1wiIiwiZmllbGRzIjpbXSwib2Zmc2V0IjowLCJ0aW1lZnJhbWUiOiI4NjQwMCIsImdyYXBobW9kZSI6ImNvdW50IiwidGltZSI6eyJ1c2VyX2ludGVydmFsIjowfSwic3RhbXAiOjE0MzgxODk0NTcwMzd9 17:04:38 <bauzas> but we can move on 17:04:48 <alaski> okay. yeah, if not directly cells related 17:05:03 <alaski> melwitt: thanks for the update, glad it's good 17:05:09 <alaski> #topic Open discussion 17:05:45 <alaski> just to update everyone, on my end I'm digging into the request spec stuff 17:06:05 <alaski> I reworked my patch series to split it along different lines which took a bit, and bauzas reworking things hasn't helped :) 17:06:14 <alaski> working on testing now and then I shoudl get some stuff up 17:06:33 <belmoreira> thanks for the reviews on https://review.openstack.org/#/c/201606/ 17:06:37 <bauzas> alaski: yeah, sorry about that, but I had some good reviews :) 17:06:37 <melwitt> cool 17:06:49 <bauzas> alaski: so I had to work on some new PSs 17:07:03 <bauzas> alaski: that's the main problem when you have long series :) 17:07:24 <alaski> belmoreira: np 17:07:42 <alaski> bauzas: yeah, definitely understand. just makes my job harder :) 17:07:57 <belmoreira> alaski: I have been travelling but should be good to continue the work now 17:08:16 <bauzas> alaski: eh eh, someone said in the midcycle that he was not happy with some SchedulerHints object so I had to rebase my whole series :p 17:08:38 <alaski> belmoreira: great. that will be helpful 17:08:52 <alaski> bauzas: I wouldn't listen to that guy 17:09:11 <alaski> that does simplify things a bit though, so it's actually good 17:09:35 <bauzas> no worries :) 17:10:04 <alaski> anyone else have a topic today? 17:10:21 <alaski> until the summit we're likely to have a light agenda of just updates in this meeting 17:10:23 <bauzas> nothing apart saying I should be off for the next 2 weeks 17:10:32 <alaski> again? :) 17:10:40 <belmoreira> I have a question 17:10:42 <bauzas> I'm French, remember ? :) 17:11:09 <belmoreira> we are populating instance_types with default flavors 17:11:23 <bauzas> belmoreira: in https://review.openstack.org/#/c/201606/ ? 17:11:44 <alaski> oh, right. that's a terrible bit of code somewhere 17:11:51 <belmoreira> should we continue with this in flavors table? 17:12:01 <belmoreira> bauzas: is not there 17:12:44 <alaski> belmoreira: I would like to say no. it's a bad thing we do currently and we should stop. but I don't know all the implications of that so we should probably bring this up at the nova meeting 17:12:52 <bauzas> belmoreira: alaski: oh ok, any code I could look at ? 17:13:21 <bauzas> oh, you mean when instanciating the instance_types table? 17:13:31 <alaski> bauzas: I think it's in the 216_havana.py migration 17:13:32 <bauzas> isn't that a devstack-ish thing ? 17:13:39 <belmoreira> I would be surprised if people are relying in those flavors 17:13:45 <bauzas> oh right 17:14:16 <bauzas> alaski: yep _populate_instance_types in 216_havana.py, correct 17:14:19 <alaski> belmoreira: me too. but I'm often suprised 17:14:33 <bauzas> well, small clouds do 17:14:42 <melwitt> migration adds some flavors? that seems odd 17:14:43 <bauzas> as an operator, I was relying on those :) 17:14:56 <bauzas> melwitt: yeah, was thinking it was devstack or something else 17:15:01 <alaski> melwitt: exactly. we bootstrap nova with some flavors 17:15:15 <melwitt> bauzas: yeah, I thought the same until alaski said it 17:15:27 <alaski> bauzas: devstack would be the right place to do it, or maybe a nova-manage command though I don't really like that 17:15:42 <bauzas> alaski: devstack and/or package vendors 17:15:47 <alaski> it feels like something we should just document, "add these flavors to get started" 17:16:05 <bauzas> that's just what we do for all the other data :) 17:16:13 <belmoreira> alaski: +1 17:16:26 * bauzas remembers the early ages of deploying Diablo's debs :) 17:16:30 <vineetmenon> alaski: +! 17:16:56 <melwitt> alaski: yeah, nova-manage would have been as invasive as I'd think to go 17:17:24 <bauzas> melwitt: I would -1 a nova-manage hydrate-flavors 17:17:53 <bauzas> nova-manage is for providing something we can't do or won't do at the REST API level - very specific operations 17:17:54 <alaski> okay, I'll add this to the nova meeting agenda and we can discuss it there. might need a ML post as well 17:18:25 <bauzas> I'd be surprised that operators would know about that - I would even bet that 99% of them think it's devstack 17:19:19 <alaski> it also probably means that we(rax) have some code to strip those out when we bootstrap a new cell somewhere 17:19:58 <belmoreira> bauzas: not sure... because is what they get in the DB when setup a cloud 17:20:40 <belmoreira> at CERN they are available :) ...legacy 17:21:04 <melwitt> yeah, we don't use those canned flavors either so I guess we must delete them as part of the deployment automation 17:21:46 <belmoreira> but we repopulate the tables with our flavors... 17:22:12 <alaski> this would be a great chance to fix this 17:22:35 <alaski> any other topics today? 17:23:09 <alaski> oh, I have a question. since the meeting is going to be light until around the summit does it still feel useful to do weekly right now? 17:23:27 <alaski> I'm happy to continue doing it 17:23:37 <alaski> but don't want to waste peoples time 17:23:41 <melwitt> I'm cool with it not being weekly, but either way is fine 17:25:08 <alaski> bauzas: belmoreira any thoughts? 17:25:16 <alaski> otherwise melwitts one vote wins :) 17:25:26 <bauzas> meh 17:25:36 <vineetmenon> -_- 17:25:39 <belmoreira> I may have some questions with flavors spec but I can ping you when needed 17:25:40 <bauzas> the 1700UTC is just family business 17:25:52 <bauzas> so I'd be happy to drop that one 17:25:58 <bauzas> unless urgent queries 17:26:05 <alaski> vineetmenon: you get a vote as well :) 17:26:16 <alaski> belmoreira: that works 17:26:43 <alaski> how about we meet next week and if there's no resistance we'll alternate from then until summit 17:26:54 <alaski> and drop the 1700 for now 17:27:09 <vineetmenon> 1100UTC is too late for us in India.. but I don't want to disturb your family time. 17:27:32 <bauzas> vineetmenon: that's not really a PITA for me, just a matter of arranging my stuff :) 17:27:45 <belmoreira> yeah, Vineet and Dheeraj connect from India 17:27:56 <vineetmenon> I'll read meeting log, meanwhile 17:28:32 <alaski> we can drop either one, I just went with 1700 since bauzas spoke up 17:29:00 <bauzas> don't really look at me 17:29:13 <vineetmenon> bauzas: :) 17:29:16 <bauzas> if vineetmenon can't handle the 2100 one, I'm okay with keeping the 1700 one 17:29:27 <vineetmenon> I would prefer 1700, tbh. 17:29:45 <alaski> okay. How about we reconvene in two weeks at this time? 17:29:45 <bauzas> I can hassle alaski anytime 17:29:48 <vineetmenon> 2300UTC == 0230 IST, so it becomes terrible 17:30:03 <alaski> bauzas and I overlap quite a bit anyways 17:30:58 <alaski> #info Meeting will be held every two weeks at 1700 until summit 17:31:15 <melwitt> cool 17:31:15 <alaski> should have include a timezone :( 17:31:25 <melwitt> you can do a #undo and do it again 17:31:28 <melwitt> I think 17:31:28 <alaski> #undo 17:31:28 <openstack> Removing item from minutes: <ircmeeting.items.Info object at 0xa799a10> 17:31:35 <alaski> #info Meeting will be held every two weeks at 1700UTC until summit 17:31:37 <alaski> melwitt: thanks 17:31:40 <vineetmenon> cool 17:31:44 <alaski> anything else for today? 17:32:25 <alaski> great. thanks everyone! 17:32:36 <alaski> I will go continue to chase bauzas patch series now 17:32:40 <alaski> #endmeeting