19:00:47 #startmeeting tripleo 19:00:48 Meeting started Tue Apr 7 19:00:47 2015 UTC and is due to finish in 60 minutes. The chair is slagle. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:49 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:53 The meeting name has been set to 'tripleo' 19:00:53 hi everyone 19:01:01 o/ 19:01:15 hello 19:01:48 let me remind folks in tripleo... 19:02:31 Welcome to 5am! 19:02:38 howdy tchaypo 19:02:54 let's get rolling 19:02:59 #topic agenda 19:03:00 * bugs 19:03:00 * reviews 19:03:00 * Projects needing releases 19:03:00 * CI 19:03:00 o/ 19:03:02 * Specs 19:03:08 * Meeting time 19:03:11 * open discussion 19:03:18 Aw, I broke the agenda 19:03:20 i added a meeting time agenda item 19:04:07 spredzy ! ;) 19:04:26 #topic bugs 19:04:36 i know we have a few untriaged bugs 19:04:43 let's see if we can knock those out right now 19:04:48 * bnemec just squashed one 19:04:52 #link https://bugs.launchpad.net/tripleo/ 19:04:52 #link https://bugs.launchpad.net/diskimage-builder/ 19:04:52 #link https://bugs.launchpad.net/os-refresh-config 19:04:52 #link https://bugs.launchpad.net/os-apply-config 19:04:52 #link https://bugs.launchpad.net/os-collect-config 19:04:55 #link https://bugs.launchpad.net/os-cloud-config 19:04:57 #link https://bugs.launchpad.net/os-net-config 19:05:00 #link https://bugs.launchpad.net/tuskar 19:05:02 #link https://bugs.launchpad.net/python-tuskarclient 19:06:59 i'll reassign https://bugs.launchpad.net/tripleo/+bug/1401617 to myself 19:07:01 Launchpad bug 1401617 in tripleo "Flavor based ramdisk/kernel ID is deprecated in Ironic since Juno" [Critical,In progress] - Assigned to Clint Byrum (clint-fewbar) 19:08:28 looks like all the other crits are fix committed for tripleo 19:09:12 jdob: can you triage this one? https://bugs.launchpad.net/tripleo/+bug/1439567 19:09:13 Launchpad bug 1439567 in tripleo "[tuskar] wrong URL for tuskar in default installation of TripleO " [Undecided,New] 19:09:45 ya, I think I fixed that already 19:10:03 I think I triaged all of the other untriaged bugs. 19:10:34 indeed, i don't see any others. thanks! 19:10:42 slagle, just wanted to point out there is this https://bugs.launchpad.net/tripleo/+bug/1441170 currently blocking puppet job in ci, fix should be on the way though 19:10:42 Launchpad bug 1441170 in tripleo "ironicclient "Not Acceptable" error" [Critical,Confirmed] 19:11:00 there are a bunch of crits in tripleo that are Fix Committed that need to be gone through and possible moved to Fix Released 19:11:07 i can do that 19:11:29 #action slagle review tripleo critical Fix Committed bugs to see if any have been releaesd 19:12:03 gfidente: k, thx. who should the assignee be? 19:12:32 this is the one that's going to be fixed by updating to the later delorean trunk repo? 19:12:38 yep 19:12:48 gfidente: can you link to that review in the bug? 19:12:59 dprince has a change for it yes looking for it 19:13:04 and assign to whomever opened that reivew (dprince i believe) 19:13:14 gfidente: ok, thanks! 19:13:31 any other bug business? 19:13:50 #topic reviews 19:14:18 slagle: sure, assign to me 19:14:18 #info There's a dashboard linked from https://wiki.openstack.org/wiki/TripleO#Review_team - look for "TripleO Inbox Dashboard" 19:14:21 #link http://russellbryant.net/openstack-stats/tripleo-openreviews.html 19:14:24 #link http://russellbryant.net/openstack-stats/tripleo-reviewers-30.txt 19:14:28 #link http://russellbryant.net/openstack-stats/tripleo-reviewers-90.txt 19:15:06 instead of lamenting about stats, are there any reviews that people would like to draw attention to? 19:16:03 slagle: the most pressing thing ATM is to unblock the puppet CI job 19:16:03 I have a dib bomb of a patch (sorry) 19:16:09 i see a lot of puppet/template reviews, and dib improvements that folks are working on 19:16:26 https://review.openstack.org/#/c/167003/ 19:16:37 slagle: but tthose patches have +2's on them so once we see another passing CI run we should be a go 19:16:59 Theres some other dib smaller ones that are less important, but I just wanted to point this one out since its not a small amount of effort to review 19:17:21 hmm, yea 8 new patchsets since i last reviewed it :) 19:17:25 i'll take another look myself 19:17:40 greghaynes: that is a bomb! almost 2k lines 19:17:44 there are actual tests though! so hopefully it helps 19:17:47 yeaaa 19:17:51 a lot of that is test / fixtures though 19:19:01 #info need to unblock Puppet CI job 19:19:21 greghaynes: at some point it might make sense to have simple-init be its own project and just have DIB install it? 19:19:29 #info need reviews on https://review.openstack.org/#/c/167003/ 19:19:38 Yep, it was actually, we undid that so we could make progress on it 19:19:41 agreed though 19:20:00 dprince: is anything needed to unblock the puppet ci job other than the delorean trunk repo update? 19:20:10 greghaynes: sure, we may as well increase the size of our .git directories in the meantime 19:20:32 slagle: no, just waiting to see if the latest recheck passes like the last one did 19:20:40 * dprince has hope 19:21:01 dprince: this one? https://review.openstack.org/#/c/162666/ 19:21:13 slagle: a bit of a surprise that an ironicclient bug is what bit us here 19:21:34 slagle: actually this one: https://review.openstack.org/#/c/161277/ 19:21:53 ah, k 19:21:53 slagle: although I would take either of them I'd rather upgrade to F21 at this point 19:21:59 dprince: hah, if its an issue id be fine with revisiting that before it merges 19:22:29 #info https://review.openstack.org/#/c/161277 needed to unblock Puppet CI job 19:22:42 greghaynes: not trying to nit pick you. Just wondering if DIB *is* the right place for all that code 19:24:12 ok, any other review business? 19:24:15 Yea, Id lean towards no but lets chat about that with the others who have been working on it later 19:24:36 greghaynes: sounds goo 19:24:40 greghaynes: d 19:24:41 #topic Projects needing releases 19:25:18 so, a note about releases...if you need one, ask someone to do one for you :) 19:25:27 i don't want to see folks blocked on waiting for a release 19:25:40 since i feel we've been a bit lax about doing them weekly 19:25:44 is that a new approach instead of the semi-weekly ones we've been doing? 19:25:48 jdob: no 19:25:49 nm, you just kinda answered that 19:26:00 i just want folks to know that they can ask anytime 19:26:04 gotcha 19:26:32 are there any volunteers to do them this week? 19:26:39 if not, i'll do it 19:26:48 i'll do it 19:26:52 jdob: thanks! 19:27:00 #action jdob to do releases 19:27:01 np, will release all the things 19:27:22 #topic CI 19:27:42 any other CI topic besides unblocking the puppet job? 19:28:46 * slagle assumes not 19:28:59 I need to make another dib job 19:29:12 i think it might make sense to replace the current ha job with a puppet based one once the ha is ready there 19:29:19 something to consider anyway 19:29:27 for the image format func test I mae 19:29:42 We should probably just nix the ha job until someone actually commits to making it work. 19:30:14 bnemec: yea, we could do that too 19:30:18 bnemec, indeed I agree with slagle it'd better to switch it to puppet, when puppet implementation is ready 19:30:30 bnemec: ++ 19:30:33 its useless ATM :( 19:30:39 (even though thats my baby) 19:30:46 greghaynes: that doesnt need to run as part of tripleo ci though does it? 19:31:15 slagle: no, just dib actually 19:31:22 ++ for switching the HA job to puppet 19:31:51 One thing I would like to highly recommend is that we first optimize our image building for CI a bit 19:32:23 what types of things do you have in mind? 19:32:34 Like... either we build one image with everything (saves us time), or we build one very basic image with almost nothing and install packages at deployment time 19:32:42 Either would save us significant CI time... 19:32:52 single image++ 19:33:08 I started working on this a bit but I'd like it to be in devtest_overcloud.sh so devs can use it as well 19:33:12 a thought - the distro-minimal elments weve made in dib have very good caching potential 19:33:38 greghaynes: ack 19:33:44 so, fwiw, we're using a single image in the rdo/instack workflow and it's working quite well 19:35:23 sure, for me it is about speeding up the CI runtime 19:35:40 dprince: yea i think adding the option to devtest makes sense 19:35:41 I think in CI the roles of the various nodes are not going to change a lot or frequently so single image wouldn't cause troubles 19:36:23 +1 for single image as well, sounds reasonable to me 19:36:28 So an HA job isn't really hard from a tripleo-ci standpoint. It is just a bit of configuration work. 19:36:55 The real work here is re-structuring some things in devtest_overcloud.sh so we are less prescriptive about imaage builds 19:37:30 * gfidente will jump on injecting failure in the HA job as soon as it works decently 19:37:39 And also supporting swap in default_parameters for things like Ceph, etc. 19:37:59 #idea add option to use single image to devtest_overcloud 19:38:02 gfidente: but that is the fun part! 19:38:13 #info plan is to replace current HA job with puppet based one 19:38:36 gfidente: keep in midn we can put an ha job up and make it non-voting at first 19:39:21 any other CI topics? 19:39:36 #topic Specs 19:40:30 moving on in 5... 19:40:37 #topic Meeting time 19:40:56 the main thing i wanted to cover is if the alternate meeting time is still useful to folks? 19:41:38 it's been skipped the last few times 19:41:41 slagle, glad you raised this, the alternate time still works for me, but we skipped a few instances there I think 19:42:01 ok, well, this is probably something we need to raise on the list for increased visibility 19:42:16 but if no one is attending...we probably don't need to say we have one 19:42:22 Yeah, by definition the people who need the alternate time aren't going to be here. :-) 19:42:29 or maybe we can find a better time 19:42:30 haha 19:42:37 touche bnemec 19:42:58 bnemec: well they aren't at the alternate time either :) 19:43:01 my problem is, I skipped both 19:43:09 maybe they fled earth 19:43:21 #action slagle email list about alternate meeting time 19:43:31 if we had a single timeframe, I might have tried to join that at least 19:43:53 well, let's see what feedback we get, then re-evaluate 19:44:11 #topic open discussion 19:44:33 there is a liberty summit ideas etherpad 19:44:45 #link https://etherpad.openstack.org/p/tripleo-liberty-proposed-sessions 19:44:54 crap, that's next month already 19:46:07 let's get some ideas flowing on the etherpad 19:46:48 one question: shall we switch USE_MERGEPY to 0 by default? 19:47:11 +1 19:47:25 gfidente: i'm in favor 19:47:27 I believe the mergepy templates were already moved to a deprecated folder. 19:47:41 So we should stop using them by default. 19:47:45 yeah 19:47:52 sound logic :) 19:47:53 Although we should keep a CI job running them until we can remove them completely. 19:47:54 +1 19:48:38 I'm for switching, we'll have to do the migration carefully so as not to break CI on various projects though 19:48:51 RE the alternate time 19:49:11 fwiw I believe that was originally pushed by me because I was one of a few people who’d started in .au 19:49:47 as far as i know none of us are following tripleo now (much), and in any case it stopped being useful to .au when it moved to be more .eu friendly 19:51:36 ok. well, hopefully folks will speak up on the list 19:51:53 so we can accomdate more people if they'd like to be involved in the meetings 19:52:28 if we need to pick a totally new time for the alternate meeting, we could do that 19:52:44 slagle, I remember we had a doodle with some votes which maybe is worth pointing at for people to update their vote or remove theirselves 19:53:01 gfidente: yea, i remember, i'll look it up 19:53:30 anything else to discuss? 19:53:31 (maybe a new vote, because the people who are not interested in tripleo anymore might not see the message that they should remove themselves) 19:54:43 thx everyone! 19:54:46 #endmeeting