15:04:18 <TravT> #startmeeting openstack search
15:04:19 <openstack> Meeting started Thu Jul  9 15:04:18 2015 UTC and is due to finish in 60 minutes.  The chair is TravT. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:04:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:04:24 <openstack> The meeting name has been set to 'openstack_search'
15:04:29 <nikhil_k> o/
15:04:34 <lakshmiS> o/
15:04:35 <TravT> Always fun when your internet drops right at the wrong time
15:04:55 <TravT> Courtesy meeting reminder: lakshmiS, mclaren, nikhil, rosmaita, sigmavirus24, TravT, krykowski, david-lyle, wko, kragniz, sjmc7, ekarlso
15:04:55 <lakshmiS> yeah thats what i thought..
15:05:04 <sigmavirus24> o/
15:05:11 <sjmc7> o/
15:05:32 <kragniz> o/
15:05:49 <TravT> hey all
15:05:58 <TravT> #topic status updates
15:06:16 <TravT> Just wanted to say Happy one month anniversary to us
15:06:20 <rosmaita> o/
15:06:21 <sjmc7> :)
15:06:29 <sigmavirus24> in which TravT becomes overly attached ptl
15:06:33 <TravT> Just saw today that exactly a month ago we were voted in as a project.
15:06:34 <TravT> lol
15:06:46 <sigmavirus24> uhuh, "Just saw" righttttt ;)
15:06:51 <nikhil_k> :P
15:07:06 <TravT> well, I had to help somebody submit a new governance proposal yesterday
15:07:17 <TravT> and looked back on mine to help them.
15:07:19 <TravT> :)
15:07:20 <nikhil_k> sigmavirus24: I wonder how many moments in a day you are not in a funny mood :)
15:07:45 <TravT> but seriously, it did cause me to realize that we've actually come a long ways in a month
15:07:52 <nikhil_k> in a good way :)
15:07:57 <TravT> so thanks everybody for your help so far
15:08:17 <TravT> ok, now that tears have been s
15:08:19 <TravT> hed
15:08:22 <TravT> let's move on
15:08:36 <TravT> agenda is here: https://etherpad.openstack.org/p/search-team-meeting-agenda
15:08:43 <TravT> please add anything you want to talk about
15:08:58 <TravT> #topic Midcycle update (TravT)
15:09:06 <TravT> So, we aren't going to have an official one.
15:09:12 <sjmc7> boo
15:09:35 <TravT> yeah, if we could have worked a shared one with glance and horizon, i think it could have worked.
15:09:44 <TravT> but that proved difficult.
15:10:00 <TravT> sjmc7 and I will be at the horizon one July 21 - 23rd
15:10:22 <TravT> so, if anybody has any possibility of showing up at that one, we can spend some time on searchlight
15:10:37 <TravT> i intend on discussing integration with the horizon team there.
15:10:51 <nikhil_k> I tried to figure out my travel possibility for July 21st week and it's seems very tricky at this point.
15:11:03 <nikhil_k> it*
15:11:21 <TravT> but we still can do some virtual meetup if needed.
15:11:25 <sigmavirus24> Yeah, my travel freedom isn't exactly great =/
15:11:38 <TravT> i'd like to keep our traction going.
15:11:46 <TravT> we have made good progress.
15:11:57 <TravT> even if lots of people are split focus
15:12:28 <TravT> #topic Cross Project Updates
15:12:36 <TravT> I was only able to attend part of it
15:12:44 <TravT> but there is some discussion on new API guidelines
15:12:56 <TravT> if you have time take a look
15:12:57 <TravT> http://lists.openstack.org/pipermail/openstack-dev/2015-July/068682.html
15:13:14 <TravT> nikhil_k or david-lyle, was there anything else we should share?
15:14:04 <nikhil_k> umm, we are not going to impl passthrough for any service at this point right?
15:14:16 <nikhil_k> anyone planning on any idea like htat?
15:14:21 <sjmc7> passthrough?
15:14:46 <nikhil_k> user -> SL -> my_openstack_service ?
15:15:05 <TravT> i haven't had any thoughts around doing that.
15:15:08 <sjmc7> ah. no, don't think so. like heat does?
15:15:16 <nikhil_k> yeah :)
15:15:22 <nikhil_k> then nothing from me :)
15:15:27 <TravT> except for keystone authentication (transparent)
15:15:53 <TravT> ok then on to bug reviews
15:15:58 <TravT> #topic bugs
15:16:14 <lakshmiS> Bug 1470627: Made progress in avoiding conflict with ceilometer notification listner. Will require some changes to ceilometer code too. will be talking with them
15:16:14 <openstack> bug 1470627 in OpenStack Search (Searchlight) "Searchlight does not receive notification when deployed in same env as Ceilometer" [High,In progress] https://launchpad.net/bugs/1470627 - Assigned to Lakshmi N Sampath (lakshmi-sampath)
15:16:40 <TravT> lakshmiS: please fill us in
15:17:26 <lakshmiS> Was able to read the notification twice from ceilometer and searchlight with some code changes in ceilometer too
15:17:44 <sjmc7> without creating additional topics?
15:17:46 <lakshmiS> will be talking to their team on suggestions to changing their code for listner
15:17:53 <lakshmiS> yeah on same topics
15:18:33 <sjmc7> and you're pretty sure that's the right way to go about it?
15:19:09 <lakshmiS> yeah it feels correct to me. will know more if ceilometer team has any reservations around it
15:19:21 <TravT> does this use pools?
15:19:25 <lakshmiS> yes
15:19:32 <sjmc7> ok. i know a bunch of the ceilometer guys in fort collins if you want me (or travis) to talk to them. i'm there in two weeks
15:19:46 <lakshmiS> sound good. thx
15:19:48 <TravT> i also have found a workaround that we could suggest to them. i have found it to be pretty effective.
15:19:57 <TravT> just kill ceilometer-anotification
15:20:00 <sjmc7> is it "disable ceilometer" ?:)
15:20:05 <TravT> :-)
15:20:14 <lakshmiS> yeah why didnt i think about that ;)
15:20:21 <TravT> i dunno seems obvious to me
15:20:42 <TravT> well, lakshmiS, i think you should take it directly to them.
15:20:53 <TravT> have they had their weekly meeting yet this week
15:20:54 <sjmc7> yeah. be great to get this fixed
15:21:10 <TravT> probably you should submit a bug or something first to them
15:21:13 <lakshmiS> yeah that was my plan. will do
15:21:16 <TravT> and then bring it up.
15:21:37 <sjmc7> we should parameterize the topic in searchlight, also
15:21:54 <sjmc7> currently it's hardcoded in each plugin. probably ties in with endre's blueprint about plugin config
15:22:02 <david-lyle> ceilometer is having a virtual sprint today and tomorrow, I believe
15:22:22 <david-lyle> so no weekly meeting
15:22:26 <TravT> interesting.  do you have info on that david-lyle?
15:22:31 <TravT> maybe lakshmi could just jump in there
15:22:33 <david-lyle> will find
15:23:04 <lakshmiS> thx david-lyle:
15:23:27 <TravT> lakshmiS, if you have even the beginning of some code that you could submit as work in progress to searchlight and ceilometer
15:23:33 <TravT> that would be helpful as well.
15:23:43 <lakshmiS> yeah will do
15:23:45 <david-lyle> http://lists.openstack.org/pipermail/openstack-dev/2015-July/068911.html
15:23:45 <TravT> makes it easier to discuss
15:24:20 <TravT> sjmc7, should we talk more about the topic parameterization now, later, or next meeting?
15:24:46 <sjmc7> it's not on the agenda, but i think we could agree something pretty quick
15:25:35 <TravT> let's circle back then during ekarlo's topic  if he arrives
15:25:43 <sjmc7> k
15:25:44 <lakshmiS> TravT: sjmc7: sound like topic parameterization could be included under making plugin configurable file
15:26:10 <TravT> Bug 1469813:
15:26:10 <openstack> bug 1469813 in OpenStack Search (Searchlight) "Glance images 'Public images do not have members'" [High,In progress] https://launchpad.net/bugs/1469813 - Assigned to Steve McLellan (sjmc7)
15:26:13 <sjmc7> yeah lakshmiS, endre has a BP for it already
15:26:37 <sjmc7> i have a review up for 1469813
15:26:39 <TravT> this is pretty important
15:26:52 <TravT> otherwise, no images show up in the index on fresh devstack
15:26:54 <sjmc7> sigmavirus
15:27:07 <sjmc7> gah. sigmavirus24 had one comment i addressed
15:27:10 <sjmc7> needs reviews
15:27:17 <sigmavirus24> oh I didn't notice that you addressed it
15:27:21 <sigmavirus24> let me look again
15:27:34 <TravT> sjmc7 I just tested it about 30 minutes ago
15:27:44 <TravT> and it does seem to solve the issue
15:27:51 <TravT> #link https://review.openstack.org/#/c/196754/
15:27:57 <nikhil_k> thanks TravT
15:29:02 <TravT> ok, thanks
15:29:31 <TravT> Bug 1469336:
15:29:31 <openstack> bug 1469336 in OpenStack Search (Searchlight) "Errors are returned as html" [Undecided,New] https://launchpad.net/bugs/1469336
15:29:44 <TravT> sjmc7 you filed this one and nobody has picked it up
15:29:51 <sjmc7> i filed this. it's not hugely urgent
15:29:57 <sjmc7> just looks messy
15:30:30 <TravT> yeah, i was just looking at it, thinking that we'll want a json response when using it from horizon
15:30:38 <sjmc7> right
15:30:58 <sjmc7> well, always. we'll never want html responses. it's not a big deal, needs some middleware
15:30:58 <TravT> so, if anybody can pick it up, please do
15:31:33 <TravT> any other bugs anybody wants to discuss?
15:31:47 <sjmc7> yes..
15:31:56 <sjmc7> oh wait, the notifications one we already did
15:31:59 <sjmc7> sorry. then no
15:32:12 <TravT> we'll wait for sjmc7 to make sure he's sure
15:32:18 <TravT> :P
15:32:29 <TravT> #topic blueprint review
15:32:39 <TravT> ekarlso, are you here?
15:33:18 <TravT> i think he's usually on his way home at this time of day
15:33:26 <TravT> so, maybe he'll be on shortly.
15:33:31 <TravT> since he added those
15:33:38 <TravT> sjmc7, looks like you added a few
15:33:46 <sjmc7> i did
15:34:39 <sjmc7> nova support
15:34:51 <sjmc7> https://blueprints.launchpad.net/searchlight/+spec/nova-instances-plugin
15:35:01 <sjmc7> i have a WIP patch up  (https://review.openstack.org/#/c/198852/)
15:35:14 <sjmc7> which is a pretty close copy of what we had for the summit demo
15:35:18 <sjmc7> feedback etc welcome
15:35:40 <sjmc7> stunned silence NOT welcome
15:35:41 <nikhil_k> ++
15:36:01 <TravT> rosmaita: nikhil_k:
15:36:01 * sigmavirus24 is stunned into silence
15:36:07 <sjmc7> nikhil_k, i know you have concerns about versioned objects
15:36:07 * sigmavirus24 leaves as he's no longer welcome
15:36:10 <sjmc7> i don't know what they are :)
15:36:12 <nikhil_k> I am to follow up a bit with John and TravT and I discussed to move on it further.
15:36:13 <sjmc7> come back!!
15:36:22 <sjmc7> ok
15:36:40 <sjmc7> i'll get a unit test up for it, but otherwise there's not much more i'm doing on it now
15:36:43 <TravT> sigmavirus24: you better not leave!
15:36:56 <sigmavirus24> =P
15:36:59 <nikhil_k> sigmatroll, there?
15:37:01 <nikhil_k> :P
15:37:05 <sjmc7> if there's no more on that one, the other one was https://blueprints.launchpad.net/searchlight/+spec/openstack-client-usage
15:37:09 <sjmc7> which i filed about 30 minutes ago
15:37:22 <nikhil_k> :)
15:37:23 <TravT> wait a sec
15:37:27 <nikhil_k> there's push for it
15:37:35 <nikhil_k> but I doubt if things are there yet
15:37:35 <sjmc7> ok. i shall wait
15:37:48 <TravT> on nova instances, should we look to get work in progress of it in sooner and enhance it?
15:37:51 <TravT> or wait?
15:38:29 <TravT> definitely needs to be looked at first, of course
15:38:49 <sjmc7> i'd like to get it in if we can
15:38:57 <nikhil_k> I don;t mind merging a smaller set of impl
15:38:58 <TravT> just wondering what we'll use to decide it is enough to have first merge
15:39:19 <nikhil_k> Can we list all that we need as a part of that BP?
15:39:55 <nikhil_k> is this just instances plugin?
15:40:13 <sjmc7> yeah
15:40:22 <nikhil_k> I see the notes on summit prototype and was wondering if enhancements expected were along those lines?
15:40:39 <sjmc7> i want to get the base in first
15:41:08 <TravT> i think getting a series of patches in is easier to review
15:41:13 <sjmc7> i can't remember exactly what enhancements we listed but i'd like ot get something that works in, then make it better
15:41:26 <TravT> i don't want to fall in the trap of massive patches that never merge.
15:41:32 <TravT> especially at this stage of the project
15:41:37 <sigmavirus24> Agreed with TravT
15:41:48 <nikhil_k> I am curious more about the BP, I like smaller patches
15:42:03 <sigmavirus24> TravT: it's a good thing david-lyle witnessed our interaction at the summit otherwise people might expect that we'd be the same human being
15:42:06 <sjmc7> ok. well, give whatever feedback. i'm keen to get this in so i'll make sure i put time aside
15:42:12 <TravT> lol
15:42:13 <sjmc7> :D
15:42:15 <nikhil_k> what should be plan for the BP? Entire nova DB or just a couple of tables?
15:42:28 <sjmc7> just instances
15:42:34 <nikhil_k> ok
15:42:36 <sjmc7> don't think in terms of tables :)
15:42:44 <nikhil_k> yeah, then I was on the wrong page
15:42:59 <TravT> so, then we need another bp for more than instances
15:43:18 <nikhil_k> all we need is some check on whether this works for cells (which may need some work for initial index or may be notif too)
15:43:41 <TravT> nikhil_k, uou and rosmaita are more knowledgable about nova, so we need your help on figuring out what all we should do for nova
15:43:42 <sjmc7> yeah - my nova knowledge is pretty limited
15:44:04 <nikhil_k> basically, instances can be great but we need meta and sys meta for the UX team to benefit
15:44:29 <nikhil_k> especially if ironic and magnum are to be differentiated
15:44:34 <sjmc7> i don't know what those are, but i can benefit from instances on their own :)
15:44:53 <nikhil_k> (that's the first req I know)
15:45:40 <sjmc7> ok. let's not go too far down this now, but agree that we need more BPs around this
15:45:48 <TravT> so, i think on current patch we can keep pretty simple, but still get a test up on it and quick look from nikhil_k, others.
15:45:49 <nikhil_k> cool
15:46:24 <TravT> ok...
15:46:26 <TravT> running out of time
15:46:34 <sjmc7> i had one more BP
15:46:39 <TravT> what was your other BP in agenda?
15:46:45 <sjmc7> https://blueprints.launchpad.net/searchlight/+spec/openstack-client-usage
15:47:07 <sjmc7> very briefly (again) - the way the listener uses the openstack clients needs some TLC
15:47:43 <sjmc7> to avoid reauthenticating on every event it caches authorized clients, but i'm sure someone can come up with a neater way to do it
15:48:36 <sigmavirus24> sjmc7: I'm sure the keystone team has ideas
15:48:39 <sjmc7> and i also think that what's there right now will not deal with expired tokens properly. there's a quick fix for it, but there's also probably a nicer approach
15:48:50 <nikhil_k> use keystone middleware
15:48:53 <TravT> how does this relate to the bug
15:48:57 <TravT> Bug 1471878:
15:48:57 <openstack> bug 1471878 in OpenStack Search (Searchlight) "Service client auth in listener" [Undecided,New] https://launchpad.net/bugs/1471878
15:49:04 <sigmavirus24> sjmc7: you should check out mfedosin's work in glance_store right now
15:49:07 <sjmc7> ah, yes. i thought i'd already filed it
15:49:21 <sigmavirus24> it looks to be a good lightweight solution
15:49:22 <TravT> so they are the same?
15:49:24 <sigmavirus24> goto pypa
15:49:31 <TravT> sigmarvirus24, do you have a link?
15:49:33 <sjmc7> not really TravT - the bug is to fix the current code
15:49:42 <sjmc7> the BP is to reimplement
15:50:09 <nikhil_k> yeah, that glance_store work seems a bit icky to me
15:50:43 <nikhil_k> https://github.com/openstack/keystonemiddleware is usually the best place to use cached tokens
15:50:53 <sjmc7> remember this isn't re-authing on behalf of a user, but essentially maintaining a long-running admin auth token. the summit demo code caught Unauthorized exceptions and cleared its cached clients
15:51:06 <sjmc7> the clients don't use keystonemiddle, do they?
15:51:26 <nikhil_k> not that I know of
15:51:28 <sjmc7> anyway, this can be discussed in the SL channel, don'tneed to do it here
15:51:30 <nikhil_k> but can we wrap?
15:52:48 <sjmc7> possibly. let's discuss separately after folks have a chance to digest
15:52:51 <TravT> ok, maybe we should come back on this.
15:52:55 <sjmc7> i'll fix 147878 in the meantime
15:52:57 <TravT> not sure ekarlso is back
15:53:11 <TravT> but i would like to point out that he got a designate plugin up
15:53:31 <sigmavirus24> TravT: sorry, let me get that link
15:53:44 <sigmavirus24> #link https://review.openstack.org/#/c/199049/5
15:54:09 <TravT> thanks sigmavirus24
15:54:21 <sjmc7> will take a look, thanks
15:54:29 <TravT> okay, let's get back to the talk on topics
15:54:43 <TravT> i mean topic parameterization
15:54:50 <TravT> actually...
15:55:02 <TravT> #topic open discussion
15:55:13 <TravT> ist before going to that (we can jump to room if needed)
15:55:25 <TravT> anything else anybody needs to bring up?
15:55:40 <TravT> I need to mention I'll be out of office next Wednesday, Thursday, Friday.
15:56:02 <nikhil_k> Vacation season
15:56:16 <TravT> nikhil_k, will you be around to chair the meeting?
15:56:31 <nikhil_k> TravT: will be and can do :)
15:56:35 <TravT> thanks.
15:56:42 <TravT> anything else from anybody?
15:57:01 <nikhil_k> sjmc7: let's chat on the channel about client reauth possibilites if you've time
15:57:08 <sjmc7> sure
15:57:20 <sjmc7> let me fix the currently filed bug first
15:57:29 <TravT> FYI that nikhil_k and I recorded a PTL webinar on searchlight for the foundation yesterday
15:57:31 <sjmc7> so you can see what i want to replace
15:57:40 <TravT> they are supposed to publish those out soon
15:57:49 <nikhil_k> sjmc7: awesome thanks!
15:57:54 <nikhil_k> TravT: yay!
15:58:02 <TravT> how about we take the parameterization topic over to #openstack-searchlight
15:58:04 <nikhil_k> hopefully they will be out today/tomorrow ish
15:58:33 <TravT> Thanks everybody.  See you in the room
15:58:41 <TravT> #endmeeting