15:00:30 <sjmc7> #startmeeting openstack search
15:00:32 <openstack> Meeting started Thu Nov 10 15:00:30 2016 UTC and is due to finish in 60 minutes.  The chair is sjmc7. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:33 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:36 <openstack> The meeting name has been set to 'openstack_search'
15:00:40 <TravT> o/
15:00:45 <robcresswell> o/
15:00:48 <Kevin_Zheng> 0/
15:00:48 <sjmc7> Morning all
15:00:57 <lei-zh> o/
15:01:04 <sjmc7> Afternoon/evening for some
15:02:10 <sjmc7> Give it another minute then i’ll get started. Probably not a long one today
15:02:13 * robcresswell is some
15:02:43 <TravT> is some?
15:02:56 <sjmc7> The clock change really bit me last week; it was dark absurdly early
15:03:07 <TravT> yeah, i'm still getting used to it
15:03:13 <TravT> been waking up an hour early all week
15:03:18 <sjmc7> The UK is plunged into darkness at lunchtime
15:04:14 <sjmc7> Ok, let’s get going, but before we do, let’s welcome Kevin_Zheng to the core team - you should have +2/+W privileges on the three searchlight repos now
15:04:47 <TravT> welcome, Kevin!
15:04:50 <lei-zh> congrat!
15:04:54 <sjmc7> Thanks for all your hard work on contributions and reviews over the last few months
15:05:07 <Kevin_Zheng> thanks
15:05:48 <sjmc7> Meeting agenda’s at https://etherpad.openstack.org/p/search-team-meeting-agenda
15:05:51 <sjmc7> #topic Ocata-1 milestone
15:06:12 <sjmc7> Ocata-1 has crept up on us; it’s Nov 14th-18th, so next week
15:06:28 <sjmc7> The only thing to note is that the nova spec freeze happens at that time
15:07:04 <TravT> wow, that came fast
15:07:08 <sjmc7> I’m not sure of the latest state of play with that. I did file a load of bugs agianst nova’s server notifications
15:07:23 <TravT> there was mailing list chatter
15:07:32 <TravT> with gibi and matt saying the responded to them
15:07:32 <sjmc7> Some of them they’re taking on board, some not. The upshot was that we need to look at versioned notifications
15:08:01 <sjmc7> Because it’s unlikely they’ll modify the old ones. I’ve not had a chance to do that yet, though i dont’ think adding fields would fall under the spec freeze
15:08:57 <Kevin_Zheng> I found some other issue
15:09:19 <Kevin_Zheng> there are case like this
15:10:03 <Kevin_Zheng> deleting instance will perform some direct db delete
15:10:12 <sjmc7> in nova?
15:10:17 <Kevin_Zheng> yes
15:10:28 <Kevin_Zheng> for some related db
15:10:37 <Kevin_Zheng> like servergroup
15:10:39 <lei-zh> no notification emits?
15:10:56 <Kevin_Zheng> yes because it is direct db action
15:11:16 <Kevin_Zheng> im using phone now so i can't show the code
15:11:42 <sjmc7> i can look it up if you give me a clue where to look
15:12:42 <Kevin_Zheng> you can check for api/openstack/compute/servers.py
15:12:52 <Kevin_Zheng> for delete action
15:13:24 <Kevin_Zheng> it will finally lead to delete_db_instance
15:13:34 <sjmc7> https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/servers.py#L754
15:13:59 <Kevin_Zheng> go futhur
15:14:13 <Kevin_Zheng> to compute.api
15:14:34 <sjmc7> https://github.com/openstack/nova/blob/master/nova/compute/api.py#L1675
15:15:32 <Kevin_Zheng> maybe you guys talk the next topic first
15:15:39 <sjmc7> maybe just describe the problem, can burrow into it later :)
15:15:42 <Kevin_Zheng> I change to computer
15:15:45 <sjmc7> it’s like an ant nest
15:17:12 <sjmc7> ok, maybe come back to this in a minute
15:17:38 <sjmc7> in any case, any nova specs we want we need to prod on before next week
15:17:42 <sjmc7> #topic Project Team Gathering
15:18:11 <sjmc7> The design summit has been replaced with the PTG towards the end of the ocata release, Feb 20th-24th next year
15:18:36 <sjmc7> The first two days is for horizontal teams (which would be useful to us), last two are to concentrate on vertical work
15:19:02 <sjmc7> I haven’t requested room space but if anyone is likely to be there we can sort something out closer to the time
15:19:17 <robcresswell> I'll be there :D
15:19:33 <sjmc7> I will try to get there at least for Monday & Tuesday but it’s probably 50/50 on whether I’ll be able to go
15:20:15 <sjmc7> Think it’s in Atlanta again
15:20:24 <robcresswell> yup
15:20:32 <TravT> i'm doubtful that I'll get travel approval this time around.
15:20:46 <sjmc7> Yeah, we’re a bit wait and see at the moment
15:21:02 <TravT> but, i've also said that the last couple things I've made it to
15:21:04 <TravT> so you never know
15:21:06 <lei-zh> not sure about that, it's a brand new thing
15:21:08 <sjmc7> We could buy a VW camper and do it old school
15:21:23 <TravT> yeah, you could swing by colorado on your way to georgia
15:21:28 <TravT> only like 10 hours out of the way
15:21:31 <sjmc7> lei-zh: yeah, it is, but it’s likely to replace the design summit for a lot of people
15:21:47 <sjmc7> Going to be harder to justify sending developers to the main conference for some
15:22:03 <sjmc7> It remains to be seen how that pans out
15:23:09 <robcresswell> At least in Cisco, we're way more likely get approval for PTG than Forum
15:23:21 <robcresswell> I think the Forum will be pretty developer-lite
15:23:27 <sjmc7> Yes, us too (though ‘more’ is somewhat relative)
15:23:30 <lei-zh> I know people from my team frequently attend nova mid cycle, but this PTG remains to be seen
15:23:48 <sjmc7> I think they’re looking more on this as being the design summit rather than a mid-cycle
15:24:24 <sjmc7> Some teams may also choose to hold mid-cycles closer to the beginning of releases (maybe around m-1) but again, that remains to be seen
15:24:41 <robcresswell> I think people might just hijack the forum for midcycles
15:24:46 <robcresswell> since they'll be travelling anyway
15:26:02 <sjmc7> Yeah. Anyway, if people are going then we’ll sort things out closer to the time but keep it in mind
15:26:16 <sjmc7> Hopefully robcresswell won’t be sitting in a room on his own
15:26:55 <sjmc7> Moving on...
15:26:59 <sjmc7> #topic Reviews
15:27:00 <robcresswell> fingers crossed
15:27:27 <sjmc7> Given the o-1 milestone approaches, it’d be nice to get the more meaty reviews closer to getting merged
15:27:43 <sjmc7> The ironic plugin ( https://review.openstack.org/#/c/364462/) is pretty close
15:28:01 <sjmc7> I did get ironic running just before the summit and then got distracted with a load of stuff, so i need to revisit that
15:29:10 <sjmc7> lei-zh: the pipelining one ( https://review.openstack.org/#/c/359972/) Rick and I added a load of comments
15:29:19 <sjmc7> not sure if there’s more discussion needed on that?
15:30:15 <lei-zh> nope, I've been occupied by some other stuff this week
15:30:19 <sjmc7> yeah, me too
15:30:36 <sjmc7> no rush on it. ping me if you have any questions on the comments
15:30:37 <lei-zh> will be back next week with a new update
15:30:45 <lei-zh> ok
15:30:47 <sjmc7> Ok, cool. I’ll make sure i set aside some time for it
15:31:42 <lei-zh> also I could take a look at ironic plugin, did something related to bare metal things recently
15:32:00 <sjmc7> Ok. It’s almost there, just needs a bit more tweaking, but good to get some more eyes on it
15:32:32 <sjmc7> There are also some other more minor reviews that are hopefully a bit more straightforward (TravT added  https://review.openstack.org/#/c/359972/ i think to the agenda)
15:32:43 <sjmc7> Now everyone’s recovering from travel and such, be good to get some of those moving
15:34:23 <TravT> yeah, i'm thinking we should just push that one through
15:34:37 <sjmc7> That’s why you’re not allowed to be dicatator any more
15:34:41 <Kevin_Zheng> T_T network very bad due to stupid shopping festival
15:34:50 <sjmc7> Shopping festival??
15:34:53 <sjmc7> Sounds awful
15:35:06 <Kevin_Zheng> yeah, 11.11 in china
15:35:08 <lei-zh> haha, 30 minutes later it's chinese black friday
15:35:15 <rosmaita> sounds better than the recent election festival
15:35:29 <sjmc7> Hahaha, now everyone’s got a black friday but they’re all different days
15:35:37 <Kevin_Zheng> ready to bring up the nova instance deleting thing
15:35:45 <sjmc7> yeah, now’s a good time Kevin_Zheng
15:35:55 <Kevin_Zheng> https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L1831
15:36:04 <Kevin_Zheng> it finally lead to this
15:36:17 <sjmc7> oh my
15:37:14 <Kevin_Zheng> but maybe only instancegroup has affect to us?
15:37:19 <sjmc7> ah.. instance group ?
15:37:37 <sjmc7> actually, this brings up something i’d forgotten about
15:38:15 <Kevin_Zheng> instance group is server group the db table is called this
15:38:21 <sjmc7> yeah
15:38:45 <sjmc7> there are some other places where we record information on Servers that belong ot other services, like neutron security groups
15:38:57 <Kevin_Zheng> yes
15:39:11 <sjmc7> So I think for some notifications we need a mechanism of registering that a notification may affect other resource types
15:39:29 <sjmc7> The simplest way is to have ServerGroup listen for instance.delete events
15:39:48 <sjmc7> That might be enough
15:39:48 <Kevin_Zheng> yes, I thought about that
15:40:16 <sjmc7> the handlers are just chained together anyway so it’s no different from making Server aware of ServerGroups
15:40:33 <sjmc7> and it means less logic to figure out if a resource type is enabled
15:40:51 <Kevin_Zheng> good
15:41:03 <sjmc7> I guess for the ServerGroup you’d have to do a query against elasticsearch to find matching groups
15:41:15 <Kevin_Zheng> sure
15:41:28 <sjmc7> We do that in other places
15:41:59 <sjmc7> You won’t have an accurate version to use, so it’ll just increment
15:42:27 <Kevin_Zheng> OK, I will try it tomorrow
15:42:54 <sjmc7> Ok. Please file a bug about the behavior so we can track it
15:43:07 <Kevin_Zheng> sure
15:43:18 <sjmc7> (That can wait til tomorrow too)
15:43:29 <Kevin_Zheng> :)
15:43:36 <sjmc7> Or til after the shopping festival!
15:43:51 <sjmc7> Ok, anything else on reviews?
15:44:42 <sjmc7> Last up..
15:44:45 <sjmc7> #topic Blueprints
15:44:55 <sjmc7> I had a quick look through https://blueprints.launchpad.net/searchlight this morning
15:45:56 <sjmc7> There’s nothing jumped out as needing re-prioritizing but obviously if there’s anything anyone wants looking at bring it up in the main channel. I don’t think we have a huge amount of new stuff planned for this shorter release, so there may be some bandwidth
15:46:43 <TravT> no, we didn't come up with much
15:46:57 <sjmc7> I guess that’s good, in a sense :)
15:47:10 <sjmc7> I’ll keep working with the nova team around notifications
15:47:22 <TravT> we pretty much knew that our focus would be on getting feedback from people deploying and working with nova
15:47:27 <TravT> i'm trying to find the summit etherpad
15:48:31 <Kevin_Zheng> how will we work with nova with the list thing？
15:48:48 <Kevin_Zheng> i mean what will it be？
15:49:07 <TravT> ttps://etherpad.openstack.org/p/ocata-searchlight-summit-working-sessions
15:49:15 <sjmc7> Not sure yet. They want to use searchlight for queries across multiple cells
15:49:24 <TravT> looks like pipeline, error handling, supporting plugins, supporting nova
15:49:34 <Kevin_Zheng> calling nova list api then nova will call SL or only expose SL api
15:49:44 <sjmc7> It’ll pass through if SL is enabled
15:49:55 <sjmc7> They don’t want a hard dependency
15:50:20 <Kevin_Zheng> hmm
15:51:18 <sjmc7> It’s probably a way off yet; things move slowly :)
15:51:37 <sjmc7> That was all i had on the agenda
15:51:42 <sjmc7> #topic Open discussion
15:51:48 <sjmc7> Anyone have anything else?
15:52:22 <TravT> not today
15:52:26 <TravT> thanks sjmc7
15:52:51 <Kevin_Zheng> thanks
15:52:54 <lei-zh> nope
15:53:00 <sjmc7> If that’s it, I’ll go and refuel on caffeine. Thanks everyone. Enjoy the shopping rampage!
15:53:23 <sjmc7> #endmeeting