14:01:15 #startmeeting monasca 14:01:17 Meeting started Wed Jul 26 14:01:15 2017 UTC and is due to finish in 60 minutes. The chair is rhochmuth1. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:18 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:20 The meeting name has been set to 'monasca' 14:01:29 o/ 14:01:36 hello 14:01:42 o/ 14:01:48 o/ 14:01:49 sorry, was running slightly behind this morning 14:02:27 https://etherpad.openstack.org/p/monasca-team-meeting-agenda 14:02:57 #topic Events API - separate or add to monasca-api? 14:03:13 that's leftover from last time 14:03:15 i still need to discuss with some folks here on this topic 14:03:25 i'll try and tee that up today 14:03:54 our preference would be to keep it separate, right arturb_ ? 14:04:10 yes 14:04:19 ohhh, i thought you were more interested in combining it 14:04:22 this is the best idea t keep it separate 14:04:31 what was the motiivation 14:04:40 reasons? 14:04:51 I wanted to reuse the persister for events 14:05:11 that was the starting point for the discussion 14:05:14 I already implemented oslo.policy for events-api 14:06:28 if you would like to keep teh events api separate then i don't see a reason that i need to follow-up with folks here 14:06:34 that was the original plan anyway 14:06:58 the only reason i would need to check would be to make sure that they are OK with a different architecture 14:07:15 agreed, I think it's easier to work on it separately 14:07:46 OK, i think we can all agree right now then that the events-api will remain a separate repo 14:07:52 approved! 14:08:01 +1 14:08:05 i don't get to say that very often these days 14:08:19 +1 14:08:22 we would like to add events persisting to monasca-persister though 14:08:36 i was afraid to ask 14:08:48 :) 14:09:00 OK, i'll check with team on that issue then 14:09:32 ok 14:09:41 i don't have any major arguments why that wouldn't be OK 14:09:50 there is already a change providing events to persister: https://review.openstack.org/#/c/485113/ 14:09:59 right 14:10:20 thx arturb_ 14:10:37 no problem :) 14:11:03 i guess we can close on that topic if you believe we are done with it 14:11:14 yes, thank you 14:11:24 #topic survey 14:11:34 https://www.openstack.org/user-survey 14:12:35 we would like everyone to forward the info to anyone who uses Monasca in OpenStack 14:13:13 as far as I know it used as input for Project Navigator 14:13:36 hmmm, not sure how to do that 14:13:41 how to forward on 14:13:51 we have our email list group 14:14:00 but that probably isn't very representative 14:14:06 of real users 14:14:11 what about Helion folks? 14:14:40 hmmm, it might be possible 14:14:52 that would result in potentially a lot of users 14:15:00 good idea 14:15:40 I guess Stefano Canepa could help 14:15:53 sc 14:15:57 I'm sending the survey to all the customer I know 14:16:03 thx sc 14:16:04 cool, thanks 14:16:27 even the one that are switch from Helion '( 14:16:48 sc: just got your email with the monasca stickers. thx for sending, the team liked them. 14:17:36 sc: which one's are switching from helion, and are they moving to suse? 14:17:43 or something else? 14:17:57 rhochmuth1: it took a while to get there 14:17:59 maybe that is confidential 14:18:06 rhochmuth1: SUSE and RH 14:18:21 thx 14:18:28 rhochmuth1: but at least 2 of them asked to have monasca on RH 14:19:06 sc: that's good news 14:19:53 I'm studing different ways to deploy (I have lots of code to look at thanks to witek and his colleagues :-) ) 14:21:30 yes, we have ansible roles for RedHat here https://gitlab.com/monasca-installer 14:21:43 but it was not maintained lately 14:22:44 also, it's just a mirror of private repos and submodule references do not work, so manual cloning is necessary 14:24:46 I guess, we can move on 14:26:05 sure 14:26:18 #topic Vote for presentations in Sydney 14:26:27 https://www.openstack.org/summit/sydney-2017/vote-for-speakers/ 14:27:26 we have many Monasca sessions proposed 14:27:40 everyone wants Australia trip :) 14:27:55 hmm, looks like some intersting sessions 14:28:11 yes, some are really interesting 14:29:03 yeah, nice to see some names i've never heard of before 14:29:11 if we could only get them to come to our weekly meetings 14:29:26 and start becoming more involved with development 14:29:39 rhochmuth1: you are invite to this one https://www.openstack.org/summit/sydney-2017/vote-for-speakers/#/19672 ;-) 14:30:24 sc: thanks 14:30:50 i know those folks 14:30:56 oh no 14:31:01 :) 14:32:18 #topic Release dates 14:32:56 so, i'm guessing that we are on-track 14:33:13 I have tagged python-monascaclient for final release 14:33:21 there was a question around py35 support in the monasca-api recently 14:33:32 the stable branch should be created tomorrow 14:33:52 thx witek 14:34:13 rhochmuth1: you mean failing the Ocata goal? 14:34:40 right 14:35:24 well, I guess we have to communicate that we will not manage to do it for agent and api in Ocata release 14:35:55 they want to see that we work on this and make progress 14:36:41 in pike, isn't this a requirement 14:36:58 pardon, Pike 14:37:17 it is the goal, not requirement 14:38:25 i guess we are ok then 14:38:47 #topic monasca presence in service-types-authority 14:38:59 Howdy folks, sorry to gate-crash, but I happened to notice the meeting and monasca was in my forebrain because of https://review.openstack.org/#/c/486739/ (CC: dhellmann mordred) 14:39:10 A little background: service-types-authority is a small repository that houses what's supposed to be a definitive/authoritative list of service types known to OpenStack. Its main component is this file: https://github.com/openstack/service-types-authority/blob/master/service-types.yaml 14:39:20 Consumers of this repository would be any client of the APIs listed therein. They can e.g. use it to (programmatically) find the official service type name based on any historical aliases for that service type; or to find the docs for the service. 14:39:33 While reviewing the aforementioned patch, it came out that monasca actually has three APIs; but there's only one listed in that service-types.yaml. 14:39:39 So what I think we need is for all three entries to be listed in there. 14:39:46 Now, I'm led to understand that the doc repositories for these three projects are somewhat in flux, which is fine; but wherever they settle out ought to eventually be reflected in there. 14:39:53 What would be nice is if someone from the monasca team proposed the patch to the service-types-authority repo, cause it's y'all who know what those three APIs are called, what their aliases might be, where their docs live, etc. 14:40:44 efried: thx 14:40:57 witek: is this something that you can work on? 14:41:34 yes, I'll add it 14:41:34 or is this at area that tomasz should address 14:41:42 thx witek 14:42:12 Thanks folks. Please add me to that review whenever it's up. 14:42:21 efried: thx 14:42:35 rhochmuth1: how should we name the services? 14:42:44 Find me in -dev if you have any questions 14:43:11 efried: thanks 14:43:27 monitoring 14:43:32 logging 14:43:33 events 14:44:01 i think "monitoring" covers it for now 14:44:12 Please note naming conventions/suggestions in https://github.com/openstack/service-types-authority/blob/master/README.rst 14:44:17 events is an alias for panko 14:44:25 i see 14:44:43 And yes, also keep in mind that the names need to be universal and not conflict with others. 14:45:04 monasca-log-api has monitoring-log-api registered now 14:45:39 we also have related discussion with Tomasz on that 14:45:48 https://review.openstack.org/#/c/480030/8/api-ref/source/conf.py 14:46:52 service-type should be a generic 14:46:59 my sugestion for naming would be: monitoring, monasca-log-api, monitoring-events-api 14:47:06 monasca-log-api is as a service ty pe is the same as the project 14:47:17 s/monasca-log-api/monitoring-log-api 14:47:37 Do we need -api? 14:47:56 efried: not really, I guess 14:47:56 no 14:48:27 For the existing entry, monitoring-log-api, we can move that to an alias and make the official service type `montioring-log` (or `monitoring-logging`) 14:49:01 +1 14:49:02 monitoring-logging sounds good to me 14:49:08 should we add events? 14:49:41 i types that out and it looked rather long 14:50:06 Wouldn't worry about length. 14:50:20 We have e.g. container-infrastructure-management ;-) 14:50:34 This is for programmatic consumption 14:50:39 rhochmuth1: you don't like the name, or do you mean we should wait until implementation is ready 14:50:41 ? 14:50:57 ok, let's just stick with monitoring-logging 14:51:26 So the official set would be `monitoring`, `monitoring-logging` (+alias `monitoring-log-api`), and `monitoring-events` ? 14:52:01 sounds good to me 14:52:04 +1 14:52:13 we could also wait with events until ready, if you prefer 14:52:45 Probably not a terrible idea. Not a problem to propose -events in a separate patch later. 14:53:09 waiting until it is ready sounds good to me 14:53:42 Thanks all. 14:53:45 * efried out 14:53:59 thx 14:54:21 i think we are done with the agenda for today 14:54:31 i'll check on the monasca-persister 14:54:37 to see what folks here think 14:54:40 could you look at this review? 14:54:47 https://review.openstack.org/480030 14:55:11 it's about publishing monasca-api documentation to docs.openstack.org 14:55:28 wow, you are taking care a lot of loose-end, thx 14:55:35 i'll look at it 14:55:56 rhochmuth1: thanks 14:56:26 should we end for the day then 14:56:34 i have to go vote for some presentations 14:56:39 :) 14:57:04 #endmeeting