15:01:09 <TravT> #startmeeting openstack search
15:01:11 <openstack> Meeting started Thu Mar 31 15:01:09 2016 UTC and is due to finish in 60 minutes.  The chair is TravT. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:12 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:16 <openstack> The meeting name has been set to 'openstack_search'
15:01:31 <rosmaita> o/
15:01:35 <TravT> o/
15:01:37 <yingjun> p/
15:01:37 <RickA-HP> o/
15:01:39 <yingjun> o/
15:01:43 <sjmc7> morning
15:01:43 <lei-zh> o/
15:01:45 <lakshmiS_> o/
15:01:57 <TravT> how is everybody today?
15:02:18 <RickA-HP> Fine, but windblown.
15:02:24 <stevelle> o/
15:02:32 <rosmaita> medium
15:02:32 <TravT> o/ stevelle
15:02:50 <TravT> medium... that seems pretty average
15:02:56 <stevelle> just listening for the news
15:03:02 <yingjun> nice day, gonna sleep
15:03:15 <TravT> Allright, so let's get this going so yingjun can sleep. ;)
15:03:18 <TravT> https://etherpad.openstack.org/p/search-team-meeting-agenda
15:03:32 <TravT> #topic client release
15:03:56 <TravT> i was asked to create a searchlight client release so that a stable branch could be done
15:04:03 <TravT> so, we did that
15:04:36 <TravT> 0.2.1 https://github.com/openstack/python-searchlightclient/releases
15:04:50 <TravT> it only included a couple of very minor items
15:05:07 <TravT> so, we were able to just do the .1
15:05:21 <TravT> #topic mitaka rc2
15:05:58 <TravT> we need to get rc2 tagged on the mitaka branch
15:06:31 <sjmc7> yeah, there are a couple of patches outstanding
15:06:47 <TravT> yes, they are very close i think
15:06:52 <TravT> here's what was targetted
15:06:53 <TravT> https://launchpad.net/searchlight/+milestone/mitaka-rc2
15:07:49 <TravT> https://review.openstack.org/#/c/298947/
15:08:11 <TravT> https://review.openstack.org/#/c/298315/
15:08:39 <TravT> https://review.openstack.org/#/c/298830/
15:09:08 <TravT> lakshmiS_: i just marked the functional tests as released
15:09:18 <TravT> since that got through infra yesterday, right?
15:09:19 <lakshmiS_> Ok
15:09:22 <lakshmiS_> Yes
15:09:49 <sjmc7> that first review just needs another +2
15:10:02 <TravT> yes, all of the above should be very quick
15:10:05 <sjmc7> it’s a pretty straightforward change
15:10:15 <lakshmiS_> I will take a look
15:10:28 <sjmc7> the other one is https://review.openstack.org/#/c/299733/ which fixes designate
15:10:35 <sjmc7> which is a bit meatier
15:10:58 <TravT> yes, lakshmiS_ is that one going to be ready for serious consideration today
15:11:02 <sjmc7> for bug #1550506
15:11:04 <openstack> bug 1550506 in OpenStack Search (Searchlight) "Designate record deletion using wrong event name" [High,In progress] https://launchpad.net/bugs/1550506 - Assigned to Lakshmi N Sampath (lakshmi-sampath)
15:11:04 <TravT> or do we need to bump it
15:11:18 <lakshmiS_> Yes it is done
15:11:29 <sjmc7> ok. i can test it this morning
15:11:30 <lakshmiS_> I am adding functional tests
15:11:47 <lakshmiS_> It has none to startwith
15:12:15 <sjmc7> given it doesn’t have any now i’m not going to make that a requirement to get this in
15:12:49 <TravT> ok, there was a non-started medium on there, i just bumped it to newton.
15:13:22 <sjmc7> yeah, we need to finish these off and call it
15:13:42 <TravT> so please review all of the above. once they land, sjmc7 has been cherry picking to stable/mitaka.
15:14:10 <TravT> i just bump them on through since they are cherry picks.
15:14:41 <TravT> the only other thing i'd like to do before applying rc2 tag is to run a clean devstack from stable/mitaka just to double check.
15:16:00 <TravT> so, also on that list is the horizon plugin
15:16:05 <TravT> #topic horizon plugin
15:16:16 <TravT> This is in pretty good shape.
15:16:30 <TravT> i've got one review up and one review i'm about to put up.
15:17:06 <TravT> i let doug know that i'd be putting out a release request on it shortly... but obviously will need a few eyes on it.
15:18:07 <TravT> i did want to talk about project tracking for searchlight-ui for a minute
15:18:42 <TravT> i'd rather than we just track it in the searchlight lauchpad project for now
15:18:55 <TravT> there will also be some related changes in horizon that we need during newton.
15:19:00 <sjmc7> that’s fine with me
15:19:15 <TravT> but i think our lives will be easier tracking it in the main searchlight project
15:19:28 <TravT> i've already opened a few bugs and BPs there.
15:19:41 <TravT> for the bugs, i'm tagging them with searchlight-ui
15:20:19 <TravT> okay, any objections?
15:20:24 <sjmc7> nope
15:20:34 <RickA-HP> Sounds good to me.
15:21:07 <TravT> ok
15:21:21 <TravT> #topic hypervisor notifications
15:21:24 <TravT> rosmaita:
15:21:31 <TravT> i believe this is your topic
15:21:33 <rosmaita> yep
15:21:52 <rosmaita> my action item was to see what the nova community thought of hypervisor notifications
15:22:02 <rosmaita> they are cool with it, but will require a spec
15:22:14 <TravT> great!
15:22:15 <yingjun> FYI, i created a spec in nova: https://review.openstack.org/#/c/299807/
15:22:16 <rosmaita> i can work with yingjun on that if help is needed
15:22:23 <rosmaita> already done then!
15:22:35 <sjmc7> :)
15:22:47 <yingjun> need to get some review on that:)
15:22:49 <rosmaita> ok, i will read & comment on spec
15:23:10 <TravT> i just added myself to that
15:23:13 <rosmaita> i can push at least one nova core to look, too
15:23:29 <sjmc7> i can find someone too
15:23:39 <yingjun> nice
15:23:45 <TravT> :)
15:24:22 <TravT> any other discussion needed on this one?
15:24:27 <rosmaita> that's all from me
15:24:36 <yingjun> nope
15:24:46 <TravT> #topic notifications
15:24:50 <TravT> on a related note
15:25:01 <TravT> i've started this etherpad
15:25:02 <TravT> https://etherpad.openstack.org/p/searchlight-notification-checklist
15:25:12 <TravT> very rough at the moment
15:25:28 <TravT> so, i'm hoping that you all could help us fill it in
15:25:43 <TravT> basically list of resource types
15:26:04 <TravT> known data we are missing from notifications
15:26:10 <TravT> any related bugs and bps
15:26:40 <TravT> so yingjun, adding the link to your hypervisor spec would be good.
15:27:10 <TravT> i know we are tracking things in launchpad, but was trying to find a way to consolidate the info in one place on current status.
15:27:31 <TravT> etherpad may not be the best formatting tool, but seemed like a good collaboration tool to start with.
15:31:02 <TravT> This leads into the next couple of topics
15:31:11 <TravT> newton and summit planning
15:32:28 <TravT> #topic newton and summit planning
15:32:42 <TravT> For the summit we have 3 slots.
15:32:58 <TravT> 1 fishbowl and 2 working room slots
15:33:15 <TravT> right now, they are on Thursday and do not appear to overlap with glance or horizon sessions.
15:34:10 <rosmaita> wow, how did you manage that!
15:34:13 <sjmc7> who’s gonna be at the summit?
15:35:30 <TravT> rosmaita: i sent a message to mike perez requesting no overlap
15:35:34 <TravT> and he made it happen
15:35:43 <TravT> #link https://etherpad.openstack.org/p/searchlight-newton-summit
15:36:15 <rosmaita> amazing what you can accomplish if you contact the correct people
15:36:17 <sjmc7> last time i think we used the fishbowl to see who from other projects was interested?
15:36:26 <sjmc7> and to get ideas from potential users
15:36:45 <TravT> so back to sjmc7's question
15:36:54 <TravT> anybody here going to be there?
15:36:55 <TravT> :)
15:36:58 <TravT> i will be
15:37:06 <sjmc7> :)  if not we’ll go play minigolf instead
15:37:19 <TravT> we can just advertise free beer or something
15:37:26 <david-lyle> I'm in
15:37:27 <TravT> then shut the doors and lock them
15:37:31 <david-lyle> oh wait
15:37:34 <sjmc7> “there is no punch and pie"
15:37:35 <TravT> hehe
15:37:55 <TravT> o/ david-lyle
15:38:00 <david-lyle> o/
15:38:15 <TravT> so i think we have sjmc7, david-lyle, myself.
15:38:22 <rosmaita> me too
15:38:28 <TravT> rosmaita: awesome!
15:38:39 <rosmaita> (was trying to figure out appropriate ascii emoticon, gave up)
15:38:43 <TravT> you are bringing the beer right?
15:38:54 <rosmaita> sure
15:38:56 <yingjun> i might be there:)
15:39:04 <TravT> yingjun: that'd be great!
15:39:14 <TravT> lei-zh: any chance you will be there?
15:39:21 <rosmaita> or pie ... they allow guns in the venue, not sure about alcohol
15:39:27 <lei-zh> nope:-(
15:39:29 <sjmc7> woo!
15:39:34 <sjmc7> (to pie, not lei)
15:39:35 <rosmaita> (actually not sure about the guns, either)
15:39:57 <TravT> rosmaita don't scare off everybody now!
15:40:27 <TravT> lei-zh: that's too bad, i was thinking the notification forwarding would be a good candidate topic.
15:40:29 <sjmc7> lei-zh: if there are people you work with who are going and it’d be useful for us to talk to, we can do that too
15:41:09 <lei-zh> I think yuntong will be there
15:42:01 <lei-zh> maybe he can talk about pipelne things
15:42:02 <TravT> we'll want to coordinate with the swift team as well
15:42:24 <lei-zh> pipeline, as well as the forwarding plugin
15:42:52 <TravT> ok, lei-zh, can you confirm with him and add that as a proposed topic on the etherpad?
15:42:56 <TravT> https://etherpad.openstack.org/p/searchlight-newton-summit
15:43:21 <lei-zh> ok, I'll confirm with him
15:43:30 <notmyname> looks like seachlight has 2 working sessions?
15:43:44 <TravT> notmyname: yes, 2 working sessions and one fishbowl
15:43:53 <notmyname> perhaps we can do a swift working session that doesn't conflict with those
15:44:18 <notmyname> unless it's soemthing that could be even larger as a full cross-project topic
15:44:59 <TravT> possibly.
15:45:17 <TravT> i mean possibly could be broader.
15:45:34 <TravT> notmyname: sjmc7 got in touch with somebody to learn more about the rabbit issues
15:45:40 <TravT> with ceilometer
15:45:44 <TravT> and scalability
15:46:09 <TravT> the short story is that they were feeding all the logs from openstack into rabbit
15:46:15 <TravT> that is where they ran into trouble
15:46:23 <TravT> they didn't have trouble with normal service notifications
15:46:45 <sjmc7> well, that was from monasca i think
15:47:05 <TravT> right, thanks for the clarifications sjmc7
15:47:30 <notmyname> I added a searchlight topic to swift's planning etherpad. I'd love any extra details you can provide there https://etherpad.openstack.org/p/swift-newton-summit-planning
15:47:39 <TravT> thanks notmyname
15:47:44 <sjmc7> ok. i’m planning to talk to a few more folks here before the summit too
15:47:48 <sjmc7> last few weeks have been a bit mad
15:48:05 <TravT> we have a BP to prototype the direct indexing library as well.
15:50:41 <TravT> notmyname: just added some info
15:50:45 <TravT> to your etherpad
15:50:47 <notmyname> thanks
15:51:34 <TravT> ok, next week we will need to spend more time on scheduling. so be thinking of topics and add to the etherpad
15:51:48 <TravT> #topic newton prioritization
15:52:20 <TravT> We already have a list of candidate BP's for newton
15:52:21 <TravT> https://blueprints.launchpad.net/searchlight/newton
15:52:49 <TravT> if you have BPs that you think should be on there, please propose them
15:53:04 <sjmc7> i was toying with the idea of adding one for localization
15:53:22 <sjmc7> it might be a pandora’s box
15:53:25 <david-lyle> lots of highs :-)
15:53:32 <TravT> EVERYTHING IS HIGH!
15:53:37 <david-lyle> I know I am
15:53:43 <TravT> lol
15:53:46 <david-lyle> colorado and all
15:54:00 <sjmc7> every BP is just as important as the one that preceeds it, if not more so
15:54:04 <rosmaita> maybe i should bring brownies instead of pie
15:54:08 <david-lyle> sjmc7: localizing what?
15:54:20 <sjmc7> that is the question...
15:54:34 <sjmc7> localizing error messages etc is trivial
15:54:42 <sjmc7> but data, querying etc is not
15:54:47 <david-lyle> right but it's more about queries
15:54:49 <david-lyle> yes
15:55:11 <david-lyle> I think you're comparing based on unicode
15:55:12 <sjmc7> i guess naively you could set a locale for the installation and index appropriately
15:55:17 <david-lyle> and that's it, no?
15:55:19 <sjmc7> yeah - e-s has variouis language tokenizers
15:55:51 <david-lyle> but I can name an instance in russian and my colleague in german
15:56:05 * david-lyle can't name instance in russian
15:56:12 <david-lyle> but for example
15:56:16 <TravT> this might be important to consider in the cross region discussions
15:56:30 <sjmc7> yeah.. there’s a number of long and boring blog posts on this from elastic.co
15:56:40 <sjmc7> for non-analyzed stuff (most of what we have) it’s fine
15:56:53 <david-lyle> I guess status fields would be more interesting
15:56:55 <sjmc7> but for tokenizing different languages have different rules
15:57:05 <lei-zh> is this only related to full text
15:57:12 <sjmc7> largely, yes lei-zh
15:57:30 <TravT> status fields are the job of the ui to localize, i think.
15:57:34 <sjmc7> for non-analyzed fields it’s direct comparison
15:57:42 <TravT> since they are enums
15:57:43 <sjmc7> possibly, yeah. in any case, something we need to look at
15:57:52 <david-lyle> TravT: how about CLI?
15:57:55 <david-lyle> or API
15:57:55 <david-lyle> ?
15:58:07 <TravT> what does nova do for its status fields?
15:58:16 <TravT> can you request in a locale?
15:58:29 <david-lyle> I think so
15:58:42 <TravT> well, then that seems like a bug we should look at
15:58:44 * david-lyle would have to check
15:58:56 <sjmc7> ok, so i will add a BP
15:59:08 <sjmc7> at least to identify problem areas
15:59:18 <sjmc7> i’ll use my full command of foreign languages (speaking slowly and loudly)
15:59:25 <TravT> ok, that sounds fine.
15:59:44 <TravT> sjmc7 that is an excellent tactic
15:59:55 <TravT> very english speaker of you
16:00:06 <TravT> okay, we are out of time.
16:00:12 <TravT> please, look at the reviews mentioned
16:00:20 <TravT> so we can release rc2.
16:00:26 <TravT> thank you everybody!
16:00:31 <TravT> #endmeeting