02:32:24 <ekcs> #startmeeting congressteammeeting
02:32:25 <openstack> Meeting started Fri Nov 10 02:32:24 2017 UTC and is due to finish in 60 minutes.  The chair is ekcs. Information about MeetBot at http://wiki.debian.org/MeetBot.
02:32:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
02:32:28 <openstack> The meeting name has been set to 'congressteammeeting'
02:32:39 <ekcs> hi all!
02:32:53 <ekcs> topics scratch pad here as usual: https://etherpad.openstack.org/p/congress-meeting-topics
02:33:06 <ekcs> pls read/comment/add
02:33:30 <ramineni_> ekcs: hi
02:34:07 <ekcs> hi ramineni_ !
02:34:32 <ekcs> so masahito won’t be here today because of summit travels. same with last week.
02:35:00 <ekcs> not a whole lot of topics on my end so may be a short meeting.
02:35:05 <ramineni_> ekcs: oohok
02:35:14 <ramineni_> ekcs: how was the summit
02:35:31 <ekcs> summit was good. I guess we can dive into that part.
02:35:37 <ekcs> #topic summit update
02:36:33 <ekcs> Summit went well from what I could tell. Project update and onboarding session helped people understand congress better and what’s going on.
02:37:19 <ramineni_> ekcs: great
02:37:31 <ekcs> the self-healing forum discussion was good too. a lot of interest there, looked like over 30 people in that discussion. very helpful group with many different projects as well as ops involved.
02:38:06 <ramineni_> ekcs: any features congress needs to add for self-healing
02:38:29 <ekcs> we decided to start the self-healing SIG with primary purpose to collect use cases, do gap analysis, and propose reference openstack self-healing architecture by leveraging the different projcets.
02:39:00 <ramineni_> ekcs: ok
02:39:33 <ekcs> I think for Congress the most important things are datasource integrations with gnocchi, mistral, vitrage, monasca. possibly in that order.
02:40:01 <ramineni_> ekcs:ok
02:41:22 <ekcs> so we might consider putting gnocchi (with ceilometer api removed) higher on our priority. we could discuss that. I’m not sure how to rank them actually in priority. depends on the use cases we come up with actually.
02:42:04 <ramineni_> ekcs: right , adding gnocchi should be easier , as it might be using similar APIś
02:42:40 <ramineni_> ekcs: yes, if we have requirement, we can add that datasource
02:43:22 <ekcs> got it. I’ll revisit the priorities list then.
02:44:00 <ekcs> that’s all I have about the summit. anything else to discuss there?
02:44:46 <ramineni_> ekcs: no, we can move on
02:44:58 <ekcs> ok then.
02:45:03 <ekcs> #topic gate status
02:45:20 <ekcs> So it looks like gate is more or less normal now.
02:45:40 <ekcs> ramineni_ and others worked on getting replicated-mysql test back working
02:46:07 <ramineni_> ekcs: im trying to get the py35 dsvm job ..green , so that we can make it voting too
02:46:09 <ekcs> very nice to be back to having some replicated test.
02:46:32 <ekcs> got it. that’s great. yup I saw the patches for that.
02:46:54 <ekcs> any more comments or announcements or discussions around gating?
02:47:26 <ramineni_> ekcs: none for now, hopefully to get the gate errors resolved soon :)
02:48:15 <ekcs> awesome. great progress already getting the replicated jobs back on after zuul transition.
02:48:37 <ekcs> and also getting py35 dsvm to this stage, past the infra stuff down the actual congress code.
02:49:15 <ekcs> ok then let’s move along for now.
02:49:28 <ekcs> #topic patches
02:49:39 <ekcs> any patches we want to discuss?
02:50:15 <ramineni_> working on the last patch of the GUI bp, found some gaps in cli , server , so posted patches for the same
02:50:58 <ekcs> Great!
02:51:01 <ekcs> So about this patch: https://review.openstack.org/#/c/518696/
02:51:21 <ramineni_> ekcs: yes
02:52:28 <ekcs> It looks good to me. I was just clarifying that the support is added to client library but not to CLI. Right?
02:52:39 <ekcs> Looks like what I said was confusing to you.
02:52:45 <ekcs> and then your reply was confusing to me too.
02:53:20 <ramineni_> ok, cli you meant only osc?
02:53:31 <ekcs> yes.
02:53:49 <ramineni_> congress clinet , changes are needed
02:53:59 <ramineni_> osc, we can support it if you want
02:54:25 <ekcs> right. my understanding it, dashboard need the client library change, hence the patch. but it leaves osc/cli unchanged.
02:55:04 <ekcs> seems like we’re on the same page.
02:55:09 <ramineni_> ekcs: not only dashboard , any application who imports congressclient needs the change
02:55:25 <ramineni_> but ya, right now, dashboard
02:55:49 <ekcs> right we’re on the same page then =)
02:56:11 <ramineni_> ekcs: do you have plans to support library APIś in OSC
02:56:18 <ramineni_> or planning to leave it for now
02:57:00 <ramineni_> right now, we have to use curl requests for any library API;s use
02:59:01 <ekcs> I don’t have any specific plans. seems like something we’d evetually want to have, but also doesn’t seem to me to be a immiment issue now so not a high priority. just my thoughts.
02:59:19 <ekcs> the spec called for osc.
02:59:28 <ekcs> so it’s part of the plan.
02:59:55 <ramineni_> ekcs: ok , its just easier to use the feature if cli is supported
02:59:56 <ramineni_> :)
03:00:34 <ramineni_> sounds good, by the way , updated the commit message as requested
03:00:50 <ekcs> i’ve just been more focused on prioritizing things that affect use cases being pursued right now. where as adding the osc/cli is more usability. so it’s not high on my personal priotity but it certainly doesn’t hurt to add!
03:00:58 <ekcs> if it’s high on your priority then add it =)
03:01:46 <ramineni_> ekcs: no, just asking .. but sure ..will add if i find time .. :)
03:01:47 <ekcs> awesome thanks!
03:02:21 <ekcs> ok maybe we can clarify this more quickly in real time: https://review.openstack.org/#/c/518697/
03:03:28 <ekcs> I haven’t looked carefully at what you’ve encountered that needs it.
03:03:31 <ramineni_> ekcs: ya, its not working without this change as we cant pass None in body here
03:03:35 <ramineni_> https://review.openstack.org/#/c/518696/1/congressclient/v1/client.py
03:03:55 <ramineni_> null is not accepted identifier in python
03:04:43 <ekcs> hmm weird so what happens if we do body=None? some kind of error?
03:04:58 <ramineni_> yes, no a vald json type or something im getting
03:05:16 <ramineni_> not**
03:06:10 <ekcs> hmm. ok got it. well I’m not well versed on API conventions and what a body should be, null or {}. if you know than that’s great. if not I’ll take a closer look at conventions.
03:06:20 <ramineni_> so instead of checking explicit None, checking both empty dict or None, both should qualify for the case
03:07:25 <ramineni_> ekcs: you can check if yo want .. if any better way is there, but for server patch , either way ..change seems valid?
03:07:47 <ramineni_> if there is empty dict , bad request error shouldnt be raised ideally
03:07:48 <ramineni_> ?
03:08:26 <ekcs> yea i guess it should be fine. i just get fussy when it comes to getting the API interface right and I’m not totally sure what’s right. i’ll take a closer look but it should be fine the way it is.
03:08:32 <ekcs> thanks for the clarifying discussion.
03:08:58 <ramineni_> ekcs: great, thanks
03:09:07 <ekcs> ok anything else to talk about on patches?
03:09:22 <ramineni_> no, done from my side
03:09:29 <ekcs> ok then.
03:09:32 <ekcs> #topic open discussion
03:09:45 <ekcs> anything else at all?
03:10:28 <ramineni_> nothing , you are back from summit, or still in sydney? no vacation plans ? :)
03:10:29 <ekcs> by the way, i’m going to be on vacation next two weeks. I’ll still be at meeting, but keep track of ML and patches. but may not be as responsive or productive. just a heads up.
03:10:47 <ramineni_> :) got it
03:11:25 <ramineni_> have a great vacation !!
03:11:26 <ekcs> I came to taiwan after sydney. just got here this a few hours ago.
03:11:43 <ramineni_> Ah , ok
03:11:55 <ekcs> thanks.
03:12:06 <ekcs> ok then. more next week?
03:12:24 <ramineni_> yes, sure,
03:12:25 <ramineni_> bye
03:12:36 <ekcs> bye!
03:12:40 <ekcs> #endmeeting