17:01:52 <dansmith> #startmeeting nova_cells
17:01:53 <openstack> Meeting started Wed Mar  1 17:01:52 2017 UTC and is due to finish in 60 minutes.  The chair is dansmith. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:01:54 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:01:57 <openstack> The meeting name has been set to 'nova_cells'
17:02:04 <dansmith> Heh, sorry, got distracted
17:02:10 <dtp> no prob
17:02:28 <melwitt> o/
17:02:42 <dansmith> #topic testing/bugs
17:02:47 <bauzas> \o
17:03:00 <dansmith> Before we get to v2, anyone aware of any failings with v1 or anything like that?
17:03:17 <dansmith> I know that mriedem fixed an issue with a re-enabled test that was breaking v1 for a bit, but that's long since fixed
17:03:20 <melwitt> tempest job failings, no
17:03:29 <melwitt> failings in general, yes
17:03:41 <dansmith> ba dum pish
17:03:54 <melwitt> haha
17:04:28 <dansmith> okay, then on to v2 stuff... I have a patch up to make devstack able to create multiple cells for the multinode jobs
17:04:37 <dansmith> https://review.openstack.org/#/c/436094/
17:04:48 <bauzas> \o/
17:04:51 <dansmith> I'm working through the failures now, but it does get pretty far
17:05:00 <dansmith> locally I can do normal stuff with a good amount of success
17:05:17 <dansmith> I have a stack of patches up to address non-cell-aware things, and a bunch of it landed last week at the ptg
17:05:24 <dansmith> https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:multi-cell-testing
17:05:48 <dansmith> So the plan is to get that fixed up enough to merge that into devstack so we're actually testing with multiple cells,
17:06:03 <dansmith> and then we can move on to improving some of the performance-y things we know we need to do
17:06:10 <dansmith> anything else on testing/bugs?
17:07:16 <dansmith> a'ight then
17:07:18 <dansmith> #topic open reviews
17:07:32 <dansmith> So obviously the above are open and *begging* for reviews
17:07:45 <dansmith> dtp: you still have your console thing open.. is that still healthy and good to look at?
17:07:53 <dtp> seems ok, yep
17:07:58 <dansmith> okay cool
17:08:24 <dansmith> melwitt: quota things
17:08:25 <dansmith> ?
17:08:48 <melwitt> same patches are still up, I have a couple of todos on the top patch still before it's ready
17:09:18 <melwitt> I'll ping people when it's done
17:09:30 <dansmith> okay what of those are dependent on placement changes?
17:09:42 <dansmith> or is that just to insulate us from cell death?
17:10:13 <melwitt> well, I had thought we would go ahead with what I have now and have a gap with a cell being down, until the placement stuff lands later this cycle
17:10:52 <dansmith> yeah, I just forgot if the placement stuff was just for the cell death part or if we'd just do that from the outset, so, cool
17:10:53 <melwitt> the todos I have are to 1) do the recheck quota after object create 2) change from counting instance.flavor.vcpus/memory_mb to instance.vcpus/memory_mb
17:10:59 <bauzas> melwitt: what is blocking you ?
17:11:26 <melwitt> bauzas: nothing blocking
17:11:52 <bauzas> Okay I thought so
17:11:57 <dansmith> any other open reviews
17:11:59 <dansmith> ?
17:13:18 <dansmith> #topic open discussion
17:13:28 <dansmith> so we had a ptg
17:13:32 <dansmith> the etherpad for that is ...
17:13:43 <dansmith> https://etherpad.openstack.org/p/nova-ptg-pike-cells
17:13:47 <dansmith> dtp: were you there?
17:13:55 <dtp> no, sorry i missed y'all
17:14:40 <dansmith> bummer
17:14:50 <dansmith> anything else for open discussion?
17:15:15 <dtp> dansmith
17:15:20 <dtp> i'll look over your changes
17:15:28 <dtp> and at some point after that i'll need a new assignment
17:15:39 <dtp> so i guess i'll just touch base w/ you and melwitt at that time
17:15:53 <dansmith> dtp: or just steal something :)
17:15:58 <dansmith> but okay, I'll try to be thinking of stuff
17:16:03 <melwitt> there's a lot of make-cells-aware stuff that needs doing right?
17:17:06 <dansmith> melwitt: potentially, yeah, but it's a bit hard to discover in parallel it seems
17:17:08 <dansmith> but yeah,
17:17:16 <melwitt> okay
17:17:31 <dansmith> dtp: if you pull down my patches, get it running, and then try to find some weirdo corner api that fails with this arrangement, that'd be a great one
17:17:49 <dtp> in devstack you mean?
17:17:51 <dansmith> dtp: poke at it until you find something that fails
17:17:51 <dansmith> yeah
17:18:03 <dtp> ok
17:18:47 <dansmith> cool, anything else?
17:19:33 <dansmith> going....
17:19:56 <dansmith> SOLD
17:19:58 <dansmith> #endmeeting