17:01:59 <mmedvede> #startmeeting third-party 17:02:01 <openstack> Meeting started Tue Feb 23 17:01:59 2016 UTC and is due to finish in 60 minutes. The chair is mmedvede. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:02:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:02:04 <openstack> The meeting name has been set to 'third_party' 17:02:46 <mmedvede> hi, anyone here for third-party working group meeing? 17:08:51 <ja3> asselin, hearing crickets? 17:09:11 <ja3> asselin_, hearing crickets? 17:09:17 <asselin_> hi...sorry I'm in the middle of some urgent issues 17:09:22 <mmedvede> ja3: hi. asselin is busy, so I am chairing today 17:09:43 <ja3> ok misha. I know the feeling, previous mtg ran over 17:10:35 <mmedvede> ok, at least there is one other person here 17:10:37 <mmedvede> :) 17:10:45 <ja3> ...and us crickets 17:11:00 <mmedvede> I have no announcements 17:11:01 <ja3> Have you been up to the usual? 17:11:13 <mmedvede> #topic CI Watch 17:11:43 <mmedvede> ja3: yes. worked a bit on ciwatch, submitted more refactoring patches 17:11:57 <mmedvede> #link Wrap db code into db.Database class https://review.openstack.org/#/c/282912/ 17:12:05 <mmedvede> #link Refactor events functions into EvenSaver class https://review.openstack.org/#/c/283692/ 17:12:44 <mmedvede> I was unable to "just" bolt on an end-to-end functional test without further refactoring things 17:13:04 <ja3> sic? not evenTsaver? 17:14:06 <mmedvede> heh, typo in commit message 17:14:15 <ja3> I haven't looked at any of your patches; still recovering from the happy surprise of having my CI x hardware arrive. 17:15:35 <mmedvede> the unit test patches are hard to review, they change a lot of things around, but should result in no functionality change 17:15:52 <mmedvede> ja3: new hardware is great! 17:16:39 <mmedvede> ok, I do not think there is anything else on ciwatch. 17:16:51 <mmedvede> #topic Common-CI Solution 17:17:30 <mmedvede> #link Zuul cannot connect to openstack gerrit: https://review.openstack.org/#/c/266041/ 17:18:44 <mmedvede> the patch above probably needs some testing. It hopefully would fix zuul connection problems for new CIs 17:18:52 <ja3> How complete is common-ci at this point, from a top > bottom standpoint? E.g. if I was constructing a dev CI cloud, is it mature enough today to function as a full stack replacement? Is it closer to 10% done or 90% done, wrt current intent? 17:19:41 <mmedvede> ja3: I would say it is much better than starting from scratch yourself. I'd say closer to 90% 17:20:08 <mmedvede> more people use it, the more kinks would be found 17:20:11 <mmedvede> and fixed 17:20:45 <mmedvede> I only use part of it, and it works well so far 17:21:04 <mmedvede> ja3: for new CIs, puppet-openstaci would be the way to start 17:21:25 <mmedvede> s/would be/is/ 17:22:18 <ja3> our "under construction" prod is still pinned to a stable branch. once we get it running on the new hw though and start to shake out voting flows, a dev flavor will be next. sounds like puppet-openstack will be dev's base then. 17:22:53 <ja3> I can't risk delaying prod voting shakeout due to anything under our control, but I have more flexibility in dev. 17:23:36 <mmedvede> ja3: what codebase are you using now? 17:24:25 <mmedvede> puppet-openstackci is nice to configure zuul / nodepool / jenkins. You would still need to deploy your cloud. 17:25:43 <mmedvede> I guess you meant stable branch of your internal configuration repository 17:26:51 <mmedvede> #topic Open Discussion 17:27:34 <mmedvede> ja3: is your CI reporting anything at the moment? 17:27:44 <ja3> misha, I've been working mostly on the tempest end, and figuring out the desired-state config of one of the test servers (since I use those for manual testing as well, so they have a longer lifetime, we configure logrotate and friends). The guy doing the puppet end of things would have to translate my "pinned to stable branch" approximation. 17:28:52 <ja3> ...basically our prod is firewalled from the puppet refactoring work until we get a dev side built and work that through the new implementation. 17:29:19 <mmedvede> this is wise 17:29:55 <ja3> ...we need to be voting more than we need to be on the latest infra code, and once we start voting (correctly) the two efforts can run in parallel 17:32:42 <mmedvede> it is a balance. The longer you wait, it could be harder to switch to using puppet-openstackci. And the more differences your setup has with upstream, the harder it is to get help from Infra team 17:34:38 <mmedvede> ja3: anything else you'd like to discuss? Seems like no one else is here 17:35:05 <ja3> fair enough. nothing to be done with it at this point I think, our priorities have consensus internally. 17:35:35 <ja3> no topics from me or the crickets I guess. 17:35:40 <mmedvede> :) 17:36:08 <mmedvede> overall goal of third-party wg is to make it easier for all CI maintainers 17:37:09 <mmedvede> and common CI helps with this by allowing to eliminate some differences 17:38:46 <mmedvede> ja3: thanks for showing up 17:38:56 <ja3> strategically, I'm all for it. from a timing standpoint, not switching horses now unless something upstream absolutely forces us to. 17:39:25 <ja3> thanks for sitting in the chair's chair, chair 17:39:46 <mmedvede> #endmeeting