19:10:12 #startmeeting tripleo 19:10:13 Meeting started Tue Aug 25 19:10:12 2015 UTC and is due to finish in 60 minutes. The chair is slagle. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:10:14 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:10:17 The meeting name has been set to 'tripleo' 19:10:54 anyone still around? :) 19:11:00 slagle: I am 19:11:07 karume: and devvesa too 19:11:09 apuimedo: hi 19:11:15 hi slagle 19:11:57 we usually run through a pre set agenda, do some bug triage, etc 19:12:13 #topic agenda 19:12:13 * bugs 19:12:13 * reviews 19:12:13 * Projects needing releases 19:12:13 * CI 19:12:15 * Specs 19:12:18 * open discussion 19:12:26 #topic bugs 19:12:43 so, i haven't done the bug triage in a while, so i suspect there are many untriaged bugs... 19:12:55 #link https://bugs.launchpad.net/tripleo/ 19:13:09 slagle: we are here from MidoNet. We'd like to integrate with Tripleo 19:13:27 I believe devvesa has been talking to some of you 19:14:07 apuimedo: indeed 19:14:15 apuimedo: there was a thread on openstack-dev i believe 19:14:48 slagle: ah, devvesa is back :-) 19:15:49 i see the filed blueprint. we don't use the blueprint workflow much, but i'm happy to go with it if it's useful to track this work 19:16:21 overall, the highlevel plan sounds reasonable and correct to me 19:16:32 slagle: how is the workflow you use? we can change it 19:16:32 did you have any followup points for discussion about it? 19:16:53 slagle: what do you use instead? 19:16:55 spec? 19:17:02 slagle: I'm afraid we still have to start :/ 19:17:46 we were using specs for a while, but they were a little heavyweight, honestly, we haven't been using anything other than tracking some work in some etherpads 19:17:51 well, we were waiting to confirm that the blueprint is okay and will be accepted 19:18:15 to kick off the work on the items that were discussed in the ml 19:18:38 apuimedo: ok i see. i think the blueprint is fine and the work will be accepted 19:18:55 i can't think of much reason anyone would really be opposed to it 19:19:14 slagle: very well, thanks 19:19:22 just for keep track... 19:19:29 #link https://blueprints.launchpad.net/tripleo/+spec/midonet-deployment-support 19:20:12 in that case, it would be useful to know the busiest hours in the irc channel and contacts for discussing? 19:21:17 it's usually around 11:00 - 16:00 UTC 19:21:46 that's the afternoon for european folks, and the morning for north america folks 19:22:16 not sure what timezone you all are in 19:22:47 same as the Brno office 19:22:53 ok 19:22:55 yep. CEST 19:23:32 looking at the blueprint...one other thing you might need is puppet module support to configure the backend 19:23:46 slagle: you mean puppet-neutron support? 19:23:56 apuimedo: yes. do you know if that's already there? 19:24:07 let me check 19:24:17 now I don't know if we have it merged already 19:24:21 yes, it is 19:24:46 it was merged during kilo lifecycle, and even backported to Juno after that 19:24:59 great 19:25:04 https://github.com/openstack/puppet-neutron/blob/master/manifests/plugins/midonet.pp 19:25:12 yea, i just saw that :) 19:25:20 so cool, that part looks done 19:26:14 we do have even the puppet module that installs the midonet SDN controller: https://forge.puppetlabs.com/midonet/midonet 19:26:18 slagle: is there any other component that has recently followed this path, so we could check how it went? 19:26:47 apuimedo: in the mail attached on the blueprint we have urls to cisco reviews 19:27:00 oh, I missed it 19:27:10 yea, the cisco n1kv support 19:27:15 https://review.openstack.org/#/c/201398/ 19:27:19 well, actually a link to a mail that has these urls 19:27:21 :) 19:27:35 that will be so helpful 19:27:45 since we are kind of noobs on tripleo 19:27:53 not kind of 19:27:56 total noobs 19:28:04 at least I am 19:28:06 ;-) 19:28:27 hahaha. yes we are, but I didn't want to seem so ignorant in front of so many people 19:29:11 we can help you all along :) 19:29:24 slagle: thanks for the pointer to that review and the help offer ;-) 19:29:49 when is the tripleo code freeze? 19:30:13 apuimedo: there isn't one 19:30:25 tripleo doesn't follow the integrated release cycle dates 19:30:36 everything is independently released 19:30:46 slagle: I see 19:31:07 so we could be "released" as soon as we are done, is that right? 19:31:15 give or take a week :P 19:31:38 apuimedo: yes, more or less. obviously there will be a code review process, etc 19:31:55 sure, I meant once that is done 19:31:57 but we can cut releases of our projects whenever we want 19:32:02 I guess there is some tagging 19:32:04 etc 19:32:14 also we should set up some kind of 'third-party' jenkins job 19:32:40 slagle: devvesa: what are teh requirements for that? 19:34:14 apuimedo: well, we've not had 3rd party ci in tripleo before, so it'd be a learnign experience for everyone :) 19:34:26 tripleo ci is a little different from other openstack ci 19:34:52 we have our own rack of servers running an openstack cloud just for tripleo 19:35:17 yes. I understand that there is no standard 'third-party' infrastructure, and we should provide a jenkins job in hour infrastructure that listens gerrit patches 19:35:23 s/hour/our 19:35:34 this is because of the requirements of tripleo to be able to create vm's with kvm virtualization, etc 19:35:40 and test pxe provisioning, etc 19:35:42 and, even it can not vote, it should be able to comment 19:35:55 slagle: it has to be hard to test 19:36:08 is the midonet code open source? 19:36:09 slagle: so how do the cisco guys aforementioned have it? 19:36:17 I thought the third party ci was a requirement 19:36:19 slagle: yes. it is 19:36:23 apuimedo: it's not a requirement 19:36:34 it's a nice to have 19:36:37 slagle: I see 19:37:01 and that is happening now on RH infra? 19:37:12 slagle: it is all open source ;-) 19:37:15 yes, it's hardware donated by red hat, running in a dmz 19:37:54 apuimedo: so, you all could either do something similar, but i have to say that is not something that is super easy to setup 19:38:05 you need hardware, etc 19:38:46 the alternative would be to have a job running on the existing rack that was just periodic, or run on demand 19:38:54 since it's all open source, that should be doable 19:38:57 slagle: not easy at all :P 19:39:07 slagle: that would be great 19:39:12 the community would have to accept it, and we'd need to verify we have capacity for it of course 19:39:16 slagle: who is managing that? 19:39:27 the drivers team? 19:39:46 apuimedo: the tripleo community manages it. derekh is one of the main guys working on it 19:40:09 i'm not saying this will definitely happen :) just an idea 19:40:13 slagle: cool 19:40:19 noted ;-) 19:40:22 i think folks could be open to it 19:40:34 we want to encourage integrations, etc 19:41:17 :-) 19:42:05 I feel encouraged already! 19:42:20 good :) 19:43:59 anything else before we wrap up? 19:44:10 feel free to reach out on #tripleo and the list 19:44:14 that's all from me ;-) 19:44:37 thank you for all the help! 19:44:37 cool, and thanks for joining 19:44:42 #endmeeting