17:00:43 <alaski> #startmeeting nova_cells
17:00:44 <openstack> Meeting started Wed Jan 20 17:00:43 2016 UTC and is due to finish in 60 minutes.  The chair is alaski. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:45 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:48 <openstack> The meeting name has been set to 'nova_cells'
17:00:59 <alaski> who's here today
17:01:02 <alaski> ?
17:01:04 <doffm> o/
17:01:06 <ccarmack> o/
17:01:06 <tonyb> o/
17:01:06 <belmoreira> o/
17:01:12 <mlavalle> o/
17:01:16 <alaski> nice
17:01:38 <mriedem> o/
17:01:39 <alaski> fair warning, I may have to disappear for a few minutes at some point.  some people are coming to fix up my house any moment now
17:01:53 <alaski> #topic Testing
17:02:11 <alaski> as far as I can tell everything is broken, but not cells related
17:02:29 <mlavalle> that's good
17:02:38 <alaski> so if nobody is aware of anything cells specific we'll move on
17:02:44 <alaski> mlavalle: half good :)
17:02:53 <alaski> #topic Open Reviews
17:02:59 <alaski> #link https://etherpad.openstack.org/p/mitaka-nova-priorities-tracking
17:03:19 <alaski> Also I just sent out http://lists.openstack.org/pipermail/openstack-dev/2016-January/084545.html which puts a story to the reviews
17:03:40 <belmoreira> considering what we discussed last meeting I updated: https://review.openstack.org/#/c/201606/
17:04:04 <alaski> belmoreira: awesome.  opened it up to take a look later
17:04:26 <mlavalle> alaski: long email
17:04:30 <alaski> #topic Open Discussion
17:04:52 <belmoreira> the soft delete is removed from the tables... the object will come when https://review.openstack.org/#/c/265282/ is ready
17:05:04 <alaski> mlavalle: yeah, and I tend to not be wordy about things.  lots in progress though
17:05:40 <mlavalle> alaski: yeah, I know you don't waste words ;-)
17:05:44 <alaski> belmoreira: great
17:05:44 <doffm> alaski: Thanks, its a good overview e-mail.
17:06:14 <ccarmack> regarding https://review.openstack.org/#/c/225199/, I chatted with andreaf, and I think his comments are satisfied.  He was going to re-review
17:06:15 <alaski> thanks
17:06:32 <alaski> ccarmack: excellent
17:06:34 <ccarmack> I'll bring it up at the qa meeting tomorrow
17:06:57 <melwitt_> alaski: I regret I won't be able to join you all at the mid cycle next week as my travel request at work wasn't approved
17:07:05 <alaski> melwitt_: :(
17:07:08 <ccarmack> and you wanted a new test that just creates a server with default sec group and ssh's into it?
17:07:36 <melwitt_> yeah :(
17:07:39 <doffm> melwitt_: Ughh. :(
17:07:39 <andreaf> ccarmack: heh I'll review that later today or tomorrow morning
17:07:52 <ccarmack> andreaf: thanks
17:08:16 <alaski> ccarmack: if we can just use test_server_basic_ops(?) that would be fine.  If not then yeah
17:08:36 <ccarmack> yea, I'm working on a new test in test_server_basic_ops
17:09:02 <alaski> melwitt_: I'm not sure if there will be a hangout option but perhaps you could join for relevant bits.  otherwise I can catch you up later
17:09:03 <ccarmack> the tricky part if the run_ssh config option which defaults to false based on tempest.conf
17:09:18 <ccarmack> sorry, run_ssh variable...
17:09:32 <melwitt_> ccarmack: I think a new test would only be needed if we couldn't get the tempest config option in
17:10:07 <ccarmack> melwitt: you mean the security_groups option?
17:10:09 <melwitt_> alaski: sure, thanks :)
17:10:09 <mlavalle> alaski: I am going to read your message to the ML carefully. Let me ask a simple question, though: are Cells V2 going to be done at the end of Mitaka?
17:10:32 <melwitt_> ccarmack: yes, the enable/disable security groups option
17:10:50 <alaski> mlavalle: no.  The aim is to have instance boot be a cellsv2 thing, and then in N we expand from there
17:11:11 <ccarmack> melwitt: I think even with the option in, all the ssh'ing tests are skipped for one reason or another
17:11:12 <mlavalle> alaski: any user level documentation yet?
17:11:32 <ccarmack> melwitt: even in the cells rc, or skipped for other reasons (eg. a bug)
17:11:53 <alaski> mlavalle: http://git.openstack.org/cgit/openstack/nova/tree/doc/source/cells.rst is all I'm aware of
17:12:15 <alaski> mlavalle: and https://review.openstack.org/#/c/267153/
17:12:24 <mlavalle> alaski: final question, how can I play with wha is available now?
17:12:29 <mlavalle> what^^^
17:12:46 <ccarmack> melwitt: and run_validation defaulting to false in tempest.conf causes test_server_basic_ops to skip the ssh
17:12:52 <melwitt__> ccarmack: I think we can enable at least a few of them by removing from the blacklist with the option
17:13:40 <ccarmack> melwitt: some of them create floating ip, although I unblacklisted test_server_basic_ops, which is ok because run_validation is false
17:13:59 <doffm> mlavalle: There isn't much to play with right now. And there isn't any operator documentation.
17:14:06 <alaski> mlavalle: most of it is in reviews still, but a normal devstack install will use the code once merged
17:14:42 <ccarmack> melwitt: so now I want to add a test to test_server_basic_ops that ssh'es into the instance, using fixed ip and default sec group
17:14:43 <alaski> ccarmack: is there an issue with turning run_validation on?
17:14:44 <mlavalle> ok, so devstack is an option
17:14:57 <mriedem> ccarmack: the nova rc has skips mostly for (1) security groups and (2) floating ips, i think
17:15:24 <mriedem> ccarmack: we can configure the cells job to not use floating IPs
17:15:24 <ccarmack> alaski, andreaf: do you know if run_validation in tempest.conf is set to false for the gate?
17:15:28 <mriedem> and only use fixed
17:15:42 <mriedem> ccarmack: we can look that up after the meeting
17:15:54 <mriedem> it's in the job definition in project-config, and defaults are in tempest.conf and devstack-gate
17:16:02 <ccarmack> mriedem: but there is a call in the tests to the floating ip client, won't that fail?
17:16:04 <alaski> ccarmack: I do not, I'm just wondering if it can be on if it's not
17:16:34 <alaski> hold on, back in a few
17:16:48 <ccarmack> once again I take over this meeting...
17:16:49 <mriedem> ccarmack: those could just be bad assumptions in tempest
17:16:50 <melwitt__> yeah. There's also one more setting I found in tempest related to fixed vs floating for ssh validation that we can set in tempest to enable more tests in conjunction with the new secgroup option
17:17:15 <mriedem> from what i recall, networking and ssh is all super complicated in tempest
17:17:28 <mriedem> they are working on a blueprint (andreaf is) for awhile to try and straighten that out
17:17:47 <mriedem> but i'm pretty sure we should be able to configure tempest to not use security groups or floating IPs (only fixed ips) and get things done
17:17:56 <ccarmack> there is the connect_method = fixed
17:18:14 <mriedem> anyway, that should probably be asked in the -qa channel after the meeting, probably need mtreinish's help
17:18:40 <mtreinish> ugh, ssh config stuff
17:19:00 <melwitt__> heh
17:19:04 <mtreinish> I really need to get andreaf or jlanoux to write a doc on how that's all supposed to be used
17:19:15 <mtreinish> because it's way too complicated, it needs to be cleaned up
17:19:47 <alaski> so it sounds like there may be a path to using test_server_basic_ops
17:20:14 <alaski> but if it just can't be done, then a new test seems like the way to go
17:20:47 <alaski> but as mriedem said, let's take that offline
17:20:54 <ccarmack> ok
17:21:22 <alaski> anything else for today?
17:21:25 <belmoreira> Do we have meeting next week?
17:21:34 <alaski> heh, I was just getting to that
17:21:42 <alaski> I'm proposing that we don't
17:22:32 <mriedem> +1
17:22:33 <mriedem> to not
17:22:35 <alaski> I think a lot of us will be together in person
17:23:04 <alaski> okay, unless someone objects we'll meet next in two weeks
17:23:34 <alaski> sounds like there's nothing else
17:23:45 <alaski> thanks for joining
17:23:52 <alaski> #endmeeting