15:00:41 #startmeeting RDO meeting (2016-01-06) 15:00:41 Meeting started Wed Jan 6 15:00:41 2016 UTC. The chair is chandankumar. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:41 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:41 The meeting name has been set to 'rdo_meeting_(2016-01-06)' 15:00:51 \o/ 15:00:52 Hullo, folks! 15:00:57 #chair rbowen 15:00:57 Current chairs: chandankumar rbowen 15:00:57 hi 15:01:03 o/ 15:01:06 #chair elmiko number80 15:01:06 Current chairs: chandankumar elmiko number80 rbowen 15:01:16 welcome to the future 15:01:24 #chair jruzicka 15:01:24 Current chairs: chandankumar elmiko jruzicka number80 rbowen 15:01:48 \o/ 15:01:53 #chair social 15:01:53 Current chairs: chandankumar elmiko jruzicka number80 rbowen social 15:02:23 #link RDO meeting etherpad link https://etherpad.openstack.org/p/RDO-Packaging 15:02:46 let me start with the first topic of 2016. 15:03:03 #topic updates on RDO packaging 15:03:30 o/ 15:03:38 rdoinfo patch for python-magnumclient is sent 15:03:42 #chair trown 15:03:42 Current chairs: chandankumar elmiko jruzicka number80 rbowen social trown 15:03:51 #link https://github.com/redhat-openstack/rdoinfo/pull/135 15:04:22 openstack-magnum package review is still in progress https://bugzilla.redhat.com/show_bug.cgi?id=1292794 15:04:27 15:04:49 o/ 15:04:56 #chair imcsk8 15:04:56 Current chairs: chandankumar elmiko imcsk8 jruzicka number80 rbowen social trown 15:05:16 number80, https://github.com/redhat-openstack/rdoinfo/pull/135 rdoinfo link 15:05:34 yup 15:05:41 just left you a comment 15:05:51 number80, will update the patchset. 15:05:55 ack 15:06:07 any more queries to RDO packaging updates? 15:06:13 yes 15:06:25 but apevec isn't here 15:06:35 I want to know status of rdoinf and it's forks 15:06:49 and whether or not the to unite the forks as one has come :) 15:06:56 *the time to unite 15:07:01 jruzicka: we're trying to have a single rdoinfo database for all branches (at least for liberty + mitaka) 15:07:11 number80, any obstacles in that? 15:07:23 jruzicka: currently, only opm has been tricky to handle 15:07:35 I planed to let rdoinfo live for a cycle or two and then apply engineering to the horrible mosnter it would become :) 15:07:36 but Alan has probably some minor tweaks in mind 15:07:46 *nods* 15:07:47 oh, astounding success 15:07:58 very well, that's all from me :) 15:08:11 jruzicka: just add the point in next week meeting agenda :) 15:08:33 jruzicka, Thanks :) 15:08:34 right 15:08:44 moving to next topic 15:08:51 #topic Python 3 porting effort 15:09:19 #action reboot py3 reviews 15:09:24 from the last meeting we have already completed oslo-libraries 15:09:27 #action number80 reboot py3 reviews 15:09:37 #link https://trello.com/c/ReowuP4z/105-python3 15:09:51 I need to do some more reviews this week 15:10:00 number80, great :) 15:10:29 number80, do we start client python 3 prting r we still wait for reviews of oslo? 15:10:37 *or 15:11:08 oslo needs to be merged before or we'll get build failures notifications from delorean 15:11:18 ok 15:11:46 any more queries related to python3 porting? 15:11:47 but well, let's already add a status point for next week, I'll try to get as much as possible merged by next week 15:12:00 ok 15:12:28 moving to next topic 15:12:35 #topic Oslo libraries updates 15:12:53 #link https://trello.com/c/xeD4dBAs/117-update-oslo-to-match-upper-constraints 15:13:55 that topic has been raised, and well, I'm in favor in upgrading oslo more frequently 15:14:23 Are you ok w/ that? suggestions? 15:14:34 i am ok with that 15:15:28 i will again start putting the list of packages version updates on etherpad so that people can take it from there 15:15:40 ok 15:15:52 update the card when you have it ready 15:16:01 number80, sure 15:16:09 moving to next topic 15:16:18 #topic Facter3 in RDO 15:16:51 I had the request, while I can update it in Fedora, CentOS has a too old version of Boost 15:17:27 Basically, I was wondering if we should target Mitaka or Liberty? 15:18:14 i guess it's safer to target Mitaka 15:18:15 due to version updates it might break some of the packages in mitaka 15:18:32 imcsk8: ok 15:18:57 number80, what is facter3? 15:19:16 chandankumar: well, it's a puppet dependency, it could break stuff for puppet stuff, but the request itself came from our puppeeters :) 15:20:01 ok 15:20:03 I'm not too far from having a facter3 build on CentOS, I need to fix my Boost package and we should be good 15:20:37 #link https://trello.com/c/EN1kbuox/116-add-facter3-in-rdo 15:20:41 #action hguemar import facter3 in Mitaka 15:21:23 I guess we can move to the next topic 15:21:46 #topic Liberty current-passed-ci repo status 15:21:56 trown, dmsimard go ahead 15:22:40 RDO Manager current-passed-ci deploys a cloud that is unable to launch instances... which is not ideal 15:22:58 except if you want a noop cloud :) 15:23:19 trown: what is actually broken? 15:23:26 I am rerunning https://ci.centos.org/view/rdo/job/rdo-delorean-promote-liberty/ with an old hash that previously had 100% tempest smoke to try to get back to a working current-passed-ci 15:23:34 ok 15:23:57 I think I found the root cause of not being able to launch instances, and filed a revert https://review.openstack.org/264160 15:24:10 "state of the art noop cloud" 15:24:21 #link https://review.openstack.org/264160 15:24:22 good 15:24:26 who does NOT want that?! 15:24:30 XD 15:24:30 hehe 15:24:37 hehe 15:24:38 however... it could be that the patch that reverts only shows an issue that existed before, because it changes vif_plugging_fatal to true 15:25:23 tldr: liberty delorean is borked, I am working on unborking it, and will send an email to the list when it is unborked 15:26:39 ok 15:26:56 trown, can we move to next topic? 15:26:59 #info liberty delorean is borked and in process to be unborked 15:27:16 perfect summary 15:27:47 chandankumar: yep 15:27:51 ok 15:28:05 #topic RDO upcoming events 15:28:09 #info Doc day - Jan 20/21 15:28:16 #link https://review.openstack.org/264160 15:28:25 o/ sorry I'm late 15:28:27 Join us here on IRC to address the open docs/website issues 15:28:28 #link https://github.com/redhat-openstack/website/issues 15:28:45 #info Mitaka test day, Jan 28/29 15:28:55 oh wow 43 closed cos issues 15:28:59 Kind of conflicts with FOSDEM, but it's what people preferred, by a narrow margin, over the alternatives. 15:29:03 *doc issues 15:29:07 And then 15:29:10 rbowen, since we have lots of events this month, we need to make promotion for each events so that we can get maximum participation 15:29:14 #info RDO Day @ FOSDEM schedule - https://www.rdoproject.org/events/rdo-day-fosdem-2016/ 15:29:16 rbowen++ 15:29:30 Yes, I'm working to get the word out, while not getting it drowned in the noise 15:29:34 January is very busy. 15:29:42 kudos to everyone working on the docs <3 15:29:44 rbowen: I just need to a good LTE connection while travelling to Brussels :) 15:29:49 There's also DevConf 15:30:07 I'll be there, but I'm not sure whether there's much RDO-related content or whatever. 15:30:13 But I know there's a lot of RDO people there. 15:30:20 I'll be presenting rdopkg on devconf 15:30:26 Awesome. 15:30:26 saturday 14:30 probably 15:30:38 thanks to number80 undying enthusiasm :) 15:30:45 jruzicka: I'll be attending! 15:30:50 hell yeah 15:30:54 In related news, if you have an upcoming event of any kind that you are attending or speaking at, I would be glad to help get the word out about it. 15:30:59 I need to get ack from aortega about FOSDEM 15:31:02 but I hope to come too 15:31:07 And, to that end, see my email to rdo-list about the OpenStack speakers burea. 15:31:09 bureau 15:31:17 That's all I have. Thank you. 15:31:23 rbowen Thanks :) 15:31:37 #topic chair for next meeting 15:31:46 I can do it I guess. 15:31:46 who wants to volunteer for next meeting? 15:31:59 #action jruzicka to chair for next meeting 15:31:59 I'm just a really bad timekeeper 15:32:15 jruzicka: one has to learn :) 15:32:21 number80, I'll do my best, sensei-sama 15:32:22 jruzicka, i will give a reminder to u 15:32:24 lol 15:32:28 roshi is better 15:32:32 *you 15:32:39 ;) 15:32:48 #topic open discussion 15:33:16 what was the first painful thing you needed to do in new year? :) 15:33:27 I'm asking because I'm wondering what to improve next. 15:33:48 Encouraging raw beginners to try out RDO as part of a test day. 15:33:50 jruzicka, i have already started going to gym from next year 15:33:50 getting out of bed 15:33:50 figuring out that liberty packstack was in fact mitaka packstack was pretty painful :) 15:34:03 sorry from this year 15:34:04 oh yeah 15:34:11 * jruzicka lol'd 15:34:33 dmsimard: still is in current-passed-ci :( 15:34:38 * imcsk8 still in bed :P 15:34:46 btw 15:34:55 I'm toying with the idea of something like toolbox.rdoproject.org 15:35:01 dmsimard, oh? 15:35:05 Ah? Tell me more. What does that mean? 15:35:06 to index all our tools and docs all over the place 15:35:15 that sounds cool 15:35:17 i.e, links to various CI environments where our jobs run on 15:35:22 Which is a bit of a condemnation of the main RDOproject.org site, hmm? 15:35:26 oh yeah, an index of all things RDO 15:35:26 like http://versiontracker.dmsimard.com/compare/liberty 15:35:33 that's a REAAALLY good idea. 15:35:37 and eventually monitoring I want to put in place 15:35:58 rbowen: it's a good point to put it on the actual website 15:35:59 dmsimard: different from the dashboard card? 15:36:07 I would love to talk more with you about the main site, and what we can do to make it useful enough that we dont need another site for that. Not that I object to the plan, but I'm frustrated with the main site and its lack of usability. 15:36:14 rbowen, I'm currently trying to make it less condemnable 15:36:15 but I'd need some hostname to host tools under (monitoring, the rpm tracker) 15:36:25 Yes, and I have hugely appreciated all of the PRs in the last few days. 15:36:33 rbowen: no, no, I agree, the index should actually go on the website 15:36:34 thanks for the quick action on them 15:36:40 but I meant a DNS for hosted tools 15:36:49 the idea is good, just integrate with the website 15:37:07 yes, we should connect all the dots 15:37:22 I'd like the version tracker to be hosted somewhere other than my personal server :P 15:37:27 many fancy stuffs by many different people 15:37:29 jruzicka, fewer dots would be appreciated I think 15:37:30 Anything I can do on the main site, I'm prepared to do, up to and includnig complete structural overhaul. 15:37:47 It's way too hard to find stuff on there now. 15:37:49 snecklifter, indeed, merging and minimising the dots is even better 15:37:52 but openstack is many things 15:37:55 with many problems 15:37:59 rbowen: I'll give it some more thought, just thinking out loud here 15:38:06 which I'm affraid will result in many tools :) 15:38:17 From a user perspective, documentation is all over the place 15:38:18 jcoufal had some great ideas a year ago, and it took so long to get the new site running I think that he moved on to other things. 15:38:34 snecklifter, really, you felt overwhelmed by documentation? 15:38:42 I've just removed a link to SpinalStack documentation over a year old 15:38:50 jruzicka, no, lost 15:39:01 snecklifter: I see where you're going but this isn't about documentation 15:39:11 so the main website heads off to tripleo for RDO-Manager 15:39:12 this is about linking to places and tools related to RDO 15:39:16 On the bright side, it's still a *tiny* site, so reorging it shouldn't be difficult, but it's still a little overwhelming. 15:39:25 which then links to fedorapeople but only for liberty 15:39:40 i.e, https://ci.centos.org/view/rdo/ & https://prod-rdojenkins.rhcloud.com/ & http://versiontracker.dmsimard.com/compare/liberty & eventually others 15:39:40 and then it was linking to old SpinalStack stuff 15:40:06 rbowen: is the way to go to help with the docs is to fix isses in here? https://github.com/redhat-openstack/website/issues 15:40:16 but I completely realise how much you guys do on very little resource 15:40:28 imcsk8: That is certainly one way, yes. 15:40:48 Although I also need to go through those carefully and ensure that they're the right things to fix. 15:41:00 dmsimard, yes, sorry, didn't mean to hijack thread :) 15:41:03 rbowen: it'll try to help with that as much is i can 15:41:15 We also need to diagram how we'd like the site laid out (information, not design) and make sure that things are a way that makes sense. 15:41:29 Which is on my list for this month, preferably before the doc day. 15:42:23 rbowen, agreed. Ideally, it would be with all the information accessible from the root node after reasonable amount of clicks :) 15:42:34 oh man I'm missing words 15:42:39 :) 15:42:42 As is the website. 15:42:44 *it would be a tree structure 15:43:00 rbowen, still better docs story then upstream :-p 15:43:18 as snecklifter pointed out, I see it as miracle we have that much docs with our manpower :) 15:43:28 The docs on rdoproject.org is just a bit awkward to browse and navigate 15:43:41 I like the way this is laid out: http://docs.openstack.org/openstack-ops/content/ 15:43:50 yes, not very consistent 15:43:57 dmsimard, I disagree, I think its fine to browse but so many dead links or old content 15:44:02 lots of different pieces of information put together 15:44:19 Perhaps a wiki would be better for collaboration, I don't know 15:44:21 rbowen: hm, seems no one did anything about the mail archives being broken yet :| still getting connection resets there, and I've tried pinging lots of people 15:44:38 but it's a minor nit, I suppose 15:44:38 Just a quick doc edit requires a lot of steps, it's a low barrier to entry, but still 15:44:45 dmsimard, lower barrier to entry to correct the problems 15:44:50 lol 15:45:12 Should people have to learn git to contribute doc edits, I'm not sure 15:45:52 dmsimard, I like git + asciidoc approach as seen here: https://www.rdoproject.org/packaging/rdo-packaging.html 15:45:55 They should be able to do wysiwyg edits directly on github. Although that's perhaps not obvious at the moment. 15:46:19 dmsimard, but that's nice for devs, not so much for users who just want to edit I guess 15:46:26 dmsimard, how about a link on the front page directing people to file an issue on github if they spot bad docs? 15:46:37 snecklifter, we also need to verify the edits made by people whether it is correct or not while using wiki 15:46:39 There is one. It's just perhaps kind of hidden. 15:46:41 like the update to the RDO-Manager one I just committed for example 15:46:46 It's way downa t the bottom of the page. 15:46:59 I don't know, I'm just trying to say that if I spot a typo, I need to fork the website on github, clone the repo, find the right page in the source, edit it, commit it, push it, submit a pull request 15:47:00 We had a "edit on github" banner for a while, but it was removed. I can't say I ever understood why. 15:47:43 I can do that 15:47:50 but perhaps some won't 15:47:53 rbowen, that would be good 15:48:38 Here: https://github.com/redhat-openstack/website/issues/244 15:48:57 In particular: https://github.com/redhat-openstack/website/pull/100 15:49:09 apevec requested that it be removed. 15:49:15 There was no discussion as to why. 15:49:37 "Fork me on github" ribbon != edit on github 15:49:52 oh right 15:49:55 yeah, "improve me on github" would be even better 15:49:58 http://aral.github.io/fork-me-on-github-retina-ribbons/right-turquoise@2x.png is kinda cheesy indeed 15:50:09 ok, I've seen "edit on github" banners on some sites. I'll see if I can find what needs done to do that. 15:50:22 rbowen: read the docs has it http://weirdo.readthedocs.org/en/latest/ 15:51:00 * number80 suggest to wrap up 15:51:15 +1 15:51:22 I have to go off to another meeting here soon. 15:51:38 number80, dmsimard will i close the meeting? 15:51:39 dmsimard: thanks. I'll see what, exactly, that links to, and replicate it on our ite. 15:51:54 s/ite/site/ 15:52:00 Sure, was just abusing the open floor :) 15:52:08 chandankumar: ^ 15:52:11 chandankumar is gonna do it! 15:52:16 nah, it's fine :) 15:52:20 HE'S GONNA DO IT 15:52:24 #endmeeting