19:03:24 <jeblair> #startmeeting infra
19:03:24 <openstack> Meeting started Tue Jul  9 19:03:24 2013 UTC.  The chair is jeblair. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:03:25 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:03:27 <openstack> The meeting name has been set to 'infra'
19:03:34 <fungi> seems sane as ever (maybe moreso)
19:03:35 <dolphm> #facepalm
19:03:44 <dhellmann> o/
19:03:45 <olaph> o/
19:03:47 <clarkb> o/
19:03:50 <zaro> o/
19:03:56 <fungi> dolphm: i have a copy of your sekrit meeting logs in my irc client anyway
19:03:58 <jeblair> dolphm: yeah, looks like you had a stray '-' in the startmeeting cmd
19:04:02 <mordred> o/
19:04:28 <anteaya> o/
19:04:32 * zul lurks
19:04:41 <jeblair> #link https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting
19:04:46 <jeblair> #link http://eavesdrop.openstack.org/meetings/infra/2013/infra.2013-07-02-19.03.html
19:04:58 <jeblair> #topic actions from last meeting
19:05:27 <pleia2> o/
19:05:49 <jeblair> so all the actions were about the quantum/neutron rename...
19:05:52 <jeblair> #topic Rename of the project formerly known as mutnuaq (mordred)
19:05:58 <fungi> well, i know action item #3 happened
19:06:01 <jeblair> so let's just jump to that
19:06:13 <jeblair> (oh, it does look like meetbot doesn't have ops)
19:06:25 <mordred> wth?
19:06:34 <jeblair> mordred: probably netsplit fallout
19:06:38 <mordred> ++
19:07:41 <fungi> he's op'd now
19:07:48 <jeblair> https://launchpad.net/~quantum-core
19:07:52 <jeblair> that doesn't exist anymore
19:07:56 <jeblair> so i assume ttx did that
19:08:14 <jeblair> #link https://wiki.openstack.org/wiki/Network/neutron-renaming
19:08:22 <mordred> https://launchpad.net/~neutron-core doens't exist either
19:08:22 <ttx> jeblair: indeed
19:08:51 <ttx> mordred: -core is maintained in gerrit now, removed the convenience almost-shadow group
19:09:04 <mordred> ttx: excellent. thank you
19:09:04 <jeblair> i think the only specifically infra related thing left is to manually publish a quantumclient
19:09:18 <fungi> if we need one going forward, it would be a ~neutron-security group or something instead
19:09:21 <jeblair> which happens after all the code rename stuff happens
19:09:30 <ttx> fungi: +1
19:09:57 <jeblair> ttx, markmcclain, mordred: know how the in-project renames are progressing?
19:10:06 <mordred> jeblair: I do not know
19:10:17 * ttx missed most of the saturday party due to some kind of family emergency, I hope it went well
19:10:32 <fungi> ttx: it went fine. we were marvellously overstaffed
19:10:55 <jeblair> ttx: i think it went very well; nothing unexpected happened, and the gating jobs worked the first time through afterwords
19:11:01 <ttx> jeblair: I've seen a few renames passing by. Will ask markmcclain during release meeting tonight
19:11:24 <jeblair> ttx: cool, thanks
19:12:11 <jeblair> any final thoughts on this topic?
19:12:17 <markmcclain> jeblair: horizon, heat, and nova left to finish
19:13:07 <jeblair> markmcclain: neato
19:13:18 <markmcclain> I do want to say thanks to everyone for their work on renaming.. you guys made it super smooth and easy
19:13:23 <jeblair> 'quantum' shows up a lot less in devstack now
19:13:51 <jeblair> markmcclain: thanks for iterating with us on the procedure; our earlier ones would not have gone as well.  :)
19:14:32 * fungi raises a toast to thorough planning
19:14:45 * jeblair cheers
19:15:16 <clarkb> it did go very smoothly. I ended up reviewing code >_>
19:15:52 <jeblair> so i think asterisk and mongodb are stale agenda items this week, or does anyone have something to say about those?
19:16:18 <fungi> i took mongo off just a moment ago
19:16:26 <fungi> meant to do it after last week's meeting
19:16:38 <jeblair> #topic open discussion
19:16:57 <fungi> looks like meetbot's oper status is good now
19:17:11 <jeblair> mordred has proposed a scope for the infra program
19:17:18 <jeblair> it's on the -dev list
19:17:24 <mordred> so far no one has disagreed
19:17:28 <mordred> which I take to be a good sign
19:17:32 <clarkb> last week was a short week was a short week as was the one before it. I don't have a whole lot to talk about. Elasticsearch is unhappy and will be for the near future but two new nodes should help make it happy again
19:17:35 * jeblair agrees
19:17:45 <fungi> i don't disagree with any of it, but didn't want to add silly +1 noise on the list
19:17:55 * mordred tried to land a pbr change for over 10 hours yesterday and failed
19:18:02 <pleia2> yeah, infra description looks good
19:18:03 <mordred> which was fun
19:18:25 <pleia2> no real updates from me, chugging along with cgit and bare metal stuff
19:18:31 <mordred> and pointed out just how much benefit I think the projected work around kicking stuff out of queue will be
19:18:43 <jeblair> mordred: what were the problems you observed?
19:18:44 <mordred> pleia2: I welcome our upcoming cgit masters
19:19:03 <jeblair> mordred: were you behind merge conflicts?
19:19:05 <mordred> jeblair: long amounts of time waiting when changes in front were unmergable
19:19:32 <mordred> jeblair: yes. most of the time I checked on the changes, they were in queue and zuul knew that their current state was pointless
19:19:46 <mordred> so I think our current ideas for improvement there would have been helpful
19:19:51 <jeblair> k.  i'm getting closer to that.
19:19:55 <ttx> mordred: is there a landing page for infra on the wiki already ?
19:20:03 <mordred> ttx: not sure
19:20:10 <mordred> ttx: I have not gotten to moving that info there
19:20:15 <fungi> ttx: does the InfraTeam page count?
19:20:19 <pleia2> we have an InfraTeam page I made, but that's more for us, not about Infrastructure
19:20:29 <fungi> i guess a separate page for the program itself is still warranted
19:20:51 <ttx> pleia2: I'd rather create an "Infrastructure" page  as a classic program landing page and point to InfraTeam from there
19:21:00 <pleia2> ttx: +1
19:21:11 <fungi> wfm
19:21:12 <jeblair> "live reload" is merged, i just wrote "deeper log directories"; next up is gearman priorities; then i think merge conflicts.
19:21:23 <jeblair> there's lots of docs about the program and team at ci.openstack.org
19:21:31 <jeblair> we should have the wiki page be short and point there
19:21:35 <mordred> yuo
19:21:36 <jeblair> to reduce redundancy, i think
19:21:38 <mordred> yup
19:21:48 <mordred> I think it should have mission statement, a link to ci.openstack.org and a link to InfraTeam
19:21:50 <mordred> and that's it
19:21:55 <clarkb> ++
19:22:07 <mordred> ttx: are you going to have a Programs page?
19:22:10 <jeblair> sounds good
19:22:21 <mordred> ttx: and/or should I  make the page Program/Infrastructure instead of just Infrastructure?
19:22:21 <ttx> mordred: there is one already
19:22:24 <fungi> mordred: agreed
19:22:46 <mordred> or https://wiki.openstack.org/wiki/Programs/Infrastructure
19:22:54 <ttx> mordred: just "Infrastructure". There is "Oslo" already" and a page for each existing project at the top-level
19:22:58 <pleia2> I'll clean up InfraTeam a bit this week, now that ci.openstack.org has been rewritten some of it is redundant
19:23:26 <ttx> mordred: that will avoid having to move all the existing ones. We'll just link it from the Programs page
19:24:21 <fungi> especially since there's also, e.g., https://wiki.openstack.org/wiki/Nova
19:24:40 <ttx> and https://wiki.openstack.org/wiki/Oslo
19:25:04 <fungi> the final iteration was that each of the server projects is also a program, right?
19:25:37 <fungi> (more or less)
19:25:45 <ttx> fungi: yes
19:25:58 <mordred> https://wiki.openstack.org/wiki/Infrastructure
19:26:05 <ttx> mordred: ok, will add link
19:26:09 <pleia2> \o/
19:27:28 <ttx> mordred: until we add some magic tagging to projects.yaml, would be great to list the repos that are under "Infra"
19:27:56 <ttx> can be as simple as openstack-infra/* + foo
19:29:09 * fungi just had a vision of a cron job or puppet exec to update some wiki stubs from a project/program registry in projects.yaml
19:29:54 <ttx> jeblair: can everyone see the -core group contents in gerrit ?
19:29:59 * dhellmann thinks if we have a bot updating a wiki we might be using the wrong tool :-)
19:30:01 <jeblair> ttx: that is the intent
19:30:03 <mordred> davidlenwell: you were working on a diagram from the bootcamp - any status on that?
19:30:09 <ttx> Because http://dowdberry.blogspot.fr/2013/07/openstack-programs-core-developers.html is wrong
19:30:12 <pleia2> oh, I did end up writing that launchpadlib script for our bug days, so it'll be much easier for me next time around
19:30:19 <fungi> dhellmann: yeah, that suggests something notwiki at that point, agreed
19:30:26 <pleia2> (it just lives on my desktop for now, if there is a useful place to put it I'm open to suggestions)
19:30:36 <ttx> jeblair: so everyone can access https://review.openstack.org/#/admin/groups/83,members for example ?
19:30:37 <davidlenwell> its still kinda messy.. I was waiting for olaph to send me his drawings.. but he seems to have been pretty busy.. I'll make it priotity to finish early this week.
19:31:05 * ttx can't tell what he is privileged for and what he isn't :)
19:31:27 <pleia2> davidlenwell: great! looking forward to seeing that :)
19:31:35 <olaph> davidlenwell: coming right up...
19:31:38 <jeblair> ttx: yep.  i think you are unpriviliged in this context (unless you are a member of a group)
19:31:40 <davidlenwell> thanks olaph
19:31:48 <jeblair> ttx: if any of them are not visible, it's a bug we can easily fix
19:31:49 <mordred> it would be neat if that link had a name like "https://review.openstack.org/#/admin/groups/nova-core,members"
19:32:02 <ttx> mordred: indeed
19:32:19 <clarkb> pleia2: I want to say jeepyb would be a good place for that maybe
19:32:49 <pleia2> clarkb: it's just a basic script that pours a bunch of stuff into stdout for me to copy into an etherpd
19:33:24 <jeblair> i think a link to github or git.o.o (when it exists) for an automatic index of the infra repos makes sense
19:33:28 <clarkb> hmm, not quite sure then. maybe just in your personal github is sufficient for now?
19:33:43 <pleia2> clarkb: yeah, that's what I'm thinking
19:33:51 <pleia2> I'll get it up later
19:36:25 <jeblair> anything else, or shall we wrap up early?
19:36:47 <mordred> wrapping up sounds good to me
19:36:59 <clarkb> yup. we have had a couple short weeks so not surprising :)
19:37:01 <fungi> ttx: for the record, if you click on the "general" link for the group there's a checkbox for "Make group visible to all registered users."
19:37:11 <ttx> fungi: cool
19:37:30 <fungi> ttx: and also a column for it in the groups list
19:37:39 <fungi> with little checkmarks
19:38:01 <ttx> jeblair, mordred: we are getting closer to h2 so please hold on non-urgent disruptive changes starting like..; thursday
19:38:41 <fungi> with a deadline like that, i will bump zuul changes to the top of my review queue
19:39:01 * mordred rushes to force-push all broken pbr changes...
19:39:06 <fungi> ha
19:39:08 <jeblair> so i'm trying to roll out a potentially disruptive change to zuul
19:39:20 <jeblair> which i wanted to do over the weekend, but was unable to
19:39:28 <jeblair> my next opportunity is wednesday afternoon
19:40:06 <ttx> jeblair: should be good, but expect heavy review activity this week
19:40:39 <mordred> jeblair: I'm going to release 0.5.18 of pbr now
19:40:47 <mordred> anybody have any concerns about that?
19:40:59 <clarkb> mordred: none from me
19:41:13 <mordred> done
19:41:25 <jeblair> well, if i break anything, i'll do the reverifies myself.  though if there are problems, it's likely to mean long gate queues
19:42:13 <fungi> if it works as intended, however, it could mean much shorter gate queues
19:42:20 <jeblair> but i think it's worth it because the next several changes to zuul are all about performance improvements
19:43:01 <mordred> +100
19:43:02 <clarkb> ++
19:43:52 <jeblair> at any rate, the change should be easy to revert, so if there's a problem that can't be immediately fixed, we'll just roll back to what we're running now
19:44:40 <mordred> sounds good to me
19:44:47 <fungi> i'm in favor
19:46:31 <davidlenwell> thanks for the scans olaph.. I'll kick this diagram out before eod tomorrow.
19:46:34 <jeblair> last call
19:48:09 <jeblair> thanks everyone!
19:48:11 <jeblair> #endmeeting