15:03:49 #startmeeting monasca 15:03:50 does it matter ? go ahead, but first maybe let's see who is here today 15:03:51 Meeting started Wed Nov 23 15:03:49 2016 UTC and is due to finish in 60 minutes. The chair is witek. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:03:52 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:03:55 The meeting name has been set to 'monasca' 15:03:59 o/ 15:04:03 o/ 15:04:04 o/ 15:04:05 o/ 15:04:13 o/ 15:05:06 #topic Reviews 15:05:20 https://review.openstack.org/#/c/395798/ - MERGED 15:05:46 rbak: it's yours 15:06:13 well, I believe since that is marged we can move forward, wasn't merged when I reposted agenda 15:06:15 You can skip it 15:06:20 thanks 15:06:33 Agent requirement out-of-sync issue https://review.openstack.org/#/c/397018/ https://review.openstack.org/#/c/315792/ 15:06:44 that's me 15:07:12 I noticed that agent won't start if following change is merged: https://review.openstack.org/#/c/315792/ 15:07:26 though something was changed in agent's requirements 'cause my change is in merge conflict 15:07:46 I did not look at it to verify it change for client will break or not the agent now 15:08:10 but the main reason to leave it here, was to discuss why agent is not hooked up in openstack requirements management process 15:08:23 yes, I wanted to ask the same 15:09:06 is anyone from HPE who could comment on that? 15:09:25 I believe the reason is given here: https://review.openstack.org/#/c/356104/ 15:09:35 there's general idea to remove oslo dependencies 15:09:54 from that point forward it would mean that controlling requirements from OS is no longer needed IMHO 15:10:23 though I was hoping for Roland or Craig to elaborate on that 15:10:51 should we move that point to next week again? 15:11:46 I think so, yes 15:11:52 ok 15:12:05 #topic Use oslo.db for sqla driver (yingjun) 15:12:59 I think the author is not there? 15:13:12 https://review.openstack.org/#/c/395897/ - this is the change 15:13:30 but I think that again we should skip that for the next week, that's rather breaking change 15:13:51 no updates on gerrit either 15:14:32 I wanted to discuss that here first before doing anything, personally I consider this good idea, but that's me 15:15:05 #topic Stop-gap measure for Grafana datasource authentication 15:15:53 rbak: you've been discussing this last week in #openstack-monasca, right? 15:16:25 Yeah 15:16:34 I think that was mostly a misunderstanding 15:17:01 There is an up to date branch of grafana with the keystone auth that they had missed 15:17:19 should we mark the blueprint then somehow? 15:17:42 Was there a blueprint with that? 15:17:50 https://blueprints.launchpad.net/monasca/+spec/grafana-auth-workarounds 15:18:03 Thanks, I'll put the comments from last weeks discussion in there 15:18:18 thanks rbak 15:18:50 #topic Postgres CLI foundation 15:19:26 the change for plugin works for mysql, so does not break anything 15:19:42 the case would be to discuss what new gates should be added 15:19:49 because there are a lot of options 15:20:00 we should configure experimental or no-vote jobs for this 15:20:11 for instance monasca-api can work with plain mysql, mysql over SQLAlchemy or postgres with SQLAlchemy 15:20:27 but by default monasca-api works with SQLAlchemy in devstack 15:20:42 so I am just bit confused about what new NV gates should be posted for monasca-api 15:20:50 I don't think we should use plain mysql 15:21:57 so the question is why is it still there if apparently it is not used, or at least not tested by CI 15:23:01 I think there was a plan to remove it from python implemenation 15:23:55 let's prepare the jobs definitions and discuss it on gerrit or next week latest 15:24:03 ok 15:24:38 #topic Publish job for monasca-grafana-datasource 15:25:26 the publish job for nodejs projects requires package.json file 15:26:04 rbak: should we create one? 15:26:32 http://lists.openstack.org/pipermail/openstack-dev/2016-November/107331.html 15:26:42 We could, although I'm still unclear why we're publishing this 15:26:59 You don't install this as a package 15:27:41 I cannot create the tag in the repo without defining the 'publish' job for the project 15:27:59 Oh, I didn't realize you couldn't even create a tag 15:28:17 Well I suppose that makes sense then 15:28:27 I did create, but the publish job failed 15:28:50 If we have the tag do we care if the publish job failed? 15:29:06 and by applying for the tag I had to specify the job :) 15:29:39 This is just a disconnect between how Grafana does things and how Openstack does things 15:29:40 I will talk with release team and ask if another option would be possible here 15:30:05 I don't mind adding the package file and publishing, it just seems unnecessary 15:30:46 ok, I'll check with release team first 15:31:10 sounds good 15:31:21 thanks 15:31:29 #topic https://storyboard.openstack.org/ - should we use that ? 15:31:53 I think we need to discuss that with Roland, so next week agenda 15:32:11 who has brought this up? 15:32:15 but in general we could use that for some bigger stories that multiple vendors might work on 15:32:16 me 15:32:37 what are the advantages? 15:32:55 as far as I understood the concept, we would create a story and tasks 15:33:13 for, for instance, events (something from the summit) 15:33:20 not only Fujitsu is interested 15:33:28 so with multiple tasks there under the stories 15:33:51 we would be able to work on changes for multiple components in parallel 15:34:18 or we could use that to discuss (as well) big features monasca would be extended with 15:34:28 could be good for coordinating the efforts 15:35:34 we should bring this up again and see if monasca in general would be interested in using that 15:35:52 I'd certainly would, but that might be tricky for other parties to move some planning here maybe 15:36:06 let's everyone build own opinion over the week and come back to this next time 15:36:08 that's why discussion is needed and checking if we can use that 15:36:11 ok 15:36:41 some else, opinions? 15:37:14 okay with that 15:37:15 has anyone used it already or knows? 15:38:00 #topic Run monasca gates on other devstack supported systems ? 15:38:09 again me 15:38:28 don't know if it makes sense, but devstack in general can be run on CentOS or Suse for instanc 15:38:53 agent (officially) is supported there (on redhat) as far as I remember 15:39:17 i was just thinking about gates that would run on something else than Ubuntu Xenial 15:39:59 Devstack attempts to support Ubuntu 14.04/16.04, Fedora 23/24, CentOS/RHEL 7, as well as Debian and OpenSUSE. 15:40:01 why would we want to add maintanace effort? what benefits do you see? 15:40:50 at Fujitsu we run things on RedHat (CentOS), we could see if everything is ok with building up python package, installing it or even integration tests itself 15:41:01 Companese often use redhat. So I think it is good idea. 15:41:24 is just thinking out loud, so it might be as well as just an idea that turned out to be wrong ;-) 15:43:13 adding RedHat support would potentially double the number of gate jobs then 15:43:33 it would, that's why I wanted to ask if that is something potentially ok with community 15:43:41 ok, sorry but I need to go now 15:43:45 i am a little bit concerned about the additional effort 15:43:46 bye 15:44:04 ok, bye Tomasz 15:44:37 shinya_kwbt: what is Companese? 15:44:53 oh, I see :) 15:44:54 Companies. 15:46:12 we're done with agenda, many points postponed 15:46:29 do you have some other topics? 15:46:50 no thx 15:48:07 ok, I close the meeting then 15:48:14 thanks bye 15:48:14 see you next week 15:48:31 thx for moderating, bye 15:48:32 thanks, bye 15:48:38 #endmeeting