19:01:42 <pabelanger> #startmeeting infra
19:01:43 <openstack> Meeting started Tue Mar  7 19:01:42 2017 UTC and is due to finish in 60 minutes.  The chair is pabelanger. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:01:44 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:01:47 <openstack> The meeting name has been set to 'infra'
19:01:51 <pabelanger> #link https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting#Agenda_for_next_meeting
19:01:57 <pabelanger> #topic Announcements
19:02:13 <pabelanger> I believe clarkb was successful in upgrading translate.o.o to xenial.
19:02:17 * mordred waves
19:02:32 <SotK> o/
19:02:38 <pabelanger> I don't think we had any outages either, so good work
19:03:11 <pabelanger> that was the only thing I had, moving on
19:03:17 <pabelanger> #topic Actions from last meeting
19:03:25 <pabelanger> #link http://eavesdrop.openstack.org/meetings/infra/2017/infra.2017-02-28-19.03.html Minutes from last meeting
19:03:33 <pabelanger> ianw launch a "beta" nodepool builder on xenial
19:03:56 <ianw> so i tried brining up a host on vexxhost as well
19:04:01 <ianw> just for something different
19:04:19 <ianw> hit some shade issues in the launch script https://review.openstack.org/440914
19:04:41 <pabelanger> #link https://review.openstack.org/#/c/440914/
19:04:47 <ianw> i also did a little hacking on the launch script
19:04:52 <ianw> #link https://review.openstack.org/440928
19:04:55 <ianw> which might be helpful
19:05:19 <ianw> but, there is no builder yet :)  will keep at it as time permits however
19:05:44 <pabelanger> understood
19:06:23 <pabelanger> ianw: did you want me to recycle the action from this week for the next meeting?
19:06:46 <ianw> pabelanger: i think just leave it with me, as it's not super high priority
19:07:08 <pabelanger> ianw: great
19:07:26 <pabelanger> if nothing else, we can move onto the next topic
19:07:46 <pabelanger> #topic Specs approval
19:07:54 <pabelanger> #link https://review.openstack.org/434951 Stackalytics Persistent Cache
19:08:05 <pabelanger> Still looking for some reviews on our stackalytics spec for mrmartin.
19:08:24 <mrmartin> thnx
19:08:35 <pabelanger> Also, does anybody know if we reached out to the stackalytics-core team for feedback?
19:08:58 <pabelanger> mrmartin: do you mind doing that?
19:09:23 <mrmartin> not yet, but I'll ping them
19:09:56 <pabelanger> great, I think fungi just wanted to give them a heads up, see what they thought
19:10:45 <pabelanger> #topic Priority Efforts
19:11:37 <pabelanger> I didn't have anything specific here. Not sure if jeblair wanted to add anything on the topic of zuulv3?
19:12:08 <jeblair> pabelanger: i wanted to discuss your tox changes, but you're the chair so i'll let you decide when to do that :)
19:12:34 <pabelanger> jeblair: sure, that's our next topic. So we can move into that.
19:12:50 <pabelanger> #topic Zuulv3 sample jobs (jeblair)
19:13:02 <pabelanger> jeblair: after you sir
19:13:24 <jeblair> pabelanger wrote some some basic jobs for zuulv3
19:13:37 <jeblair> https://review.openstack.org/#/c/438281/ and https://review.openstack.org/#/c/442180/
19:13:44 <jeblair> #link https://review.openstack.org/#/c/438281/
19:13:50 <jeblair> #link https://review.openstack.org/#/c/442180/
19:14:12 <jeblair> those are competing approaches to defining some basic tox jobs
19:14:27 <jeblair> i'd like folks who have an interest to take a look at both of them and weigh in on it
19:14:47 <pabelanger> yes, reviews welcome please.
19:14:54 <jeblair> particularly, whether one or the other is easier to follow or might be more mantainable if we use it as the base of our most common jobs in project-config
19:16:11 <Shrews> Is there a tl;dr summarizing the different approaches?
19:16:30 <Shrews> Maybe describe them in the commits?
19:16:40 <jeblair> that would probably be helpful :)
19:16:56 <pabelanger> yes, I can update them after meeting to describe the differences
19:17:01 <Shrews> thx
19:17:11 <pabelanger> for the most part, the differences are in the .zuul.yaml file
19:17:36 <jeblair> (but to summarize here, i'd say the main question is how far we should parameterize our playbooks
19:18:13 <pabelanger> ++
19:18:26 <jeblair> roles, obviously yes, but playbooks?  i think it's a legitimate question as to which is more easily understandable to casual users)
19:19:04 <ianw> https://review.openstack.org/#/c/438281/24/.zuul.yaml so the run_tox_env arg has disappared?
19:19:18 <ianw> disappeared even
19:19:49 <jeblair> ianw: yes, and been replaced by more playbooks
19:19:53 <pabelanger> ianw: right, it is not needed because we provide 3 different playbooks, which setup there own specific job vars
19:20:30 <ianw> ahh yes ok
19:20:52 <mordred> fwiw, 438281 reads nicer to me - but both are readable and understandable
19:21:03 <pabelanger> so far, I like 438281 more, only because you don't need to mental map things from .zuul.yaml, at the cost of more playbooks in tree
19:21:21 <pabelanger> but interest in others feedback
19:21:37 <pabelanger> interested*
19:23:00 <jeblair> pabelanger: maybe a ml post would be good too to collect more feedback.
19:23:26 <pabelanger> sure, I can propose something on etherpad first, then send it along
19:23:27 <mordred> pabelanger: also, it may feel better once the run-tox role fully replaces the run-tox.sh jenkins script and isn't just a passthrough
19:24:02 <jeblair> good point
19:24:16 <pabelanger> mordred: agree, for now I've been doing easy mode for our JJB to ansible conversion.
19:24:24 <mordred> pabelanger: TOTALLY - and I think the exact right call
19:24:52 <mordred> mostly just pointing out that some of the parameterization might feel artificial to folks when the role is just a shell call to a script
19:25:18 <mordred> so to sort of imagine that run-tox actually has the run-tox.sh logic in it when looking at these
19:25:56 <pabelanger> exactly
19:26:53 <pabelanger> #action pabelanger to send ML post to get more feedback on our current run-tox playbooks / role
19:28:08 <pabelanger> This was our only topic for todays meeting, so if nothing else, we can move to open discussions
19:28:47 <pabelanger> #topic Open discussion
19:30:12 <pabelanger> Seen a tweet about ubuntu 12.04 EOL next month (Apr-2017), since we still have a few servers running 12.04 we should think about a virtual sprint to fix that
19:30:30 <mordred> pabelanger: speaking of ...
19:30:36 <mordred> we have a server still running 10.04
19:30:49 <cmurphy> wow
19:30:51 <mordred> maybe it's time to turn off old-wiki?
19:31:01 <mordred> or maybe we should keep it to see how long it'll run ...
19:31:30 <pabelanger> 10.04 really?
19:31:32 <mordred> (for those who don't know, old-wiki is the very first vm that was ever spun up for openstack)
19:31:41 <mordred> pabelanger: 10.04 was very modern when that vm was created
19:31:52 <mordred> also - please don't let anyone tell you that cloud vms are not stable
19:32:08 <mordred> since that vm was created, I believe I've had 6 different laptops :)
19:32:32 <bkero> Snarf it down through wget --recursive, host that content somewhere static?
19:32:34 <pabelanger> indeed, I didn't know old-wiki.o.o was a thing
19:33:01 <mordred> oh - the content is not useful - it's all been migrated elsewhere to my knowledge
19:33:48 <pabelanger> I was planning on talking with clarkb once he is back online, maybe for next meeting has a few options for a virtual sprint to do upgrades. I don't think we have many servers that are 12.04 right now
19:33:49 <mordred> bkero: although I suppose we could also snapshot the machine image, then download it :)
19:33:52 <ianw> we could just check it's in archive.org (for posterity) and zap it?
19:34:10 <mordred> pabelanger: ++
19:34:27 <bkero> mordred: I have nothing I want on there, although if fears still exist about losing data, ianw or my option should alleviate those.
19:35:51 <pabelanger> sounds like we might be able to end early today
19:36:12 <mordred> oh wow - this is in my homedir: nova_0.9.0-0ubuntu0ppa1~lucid1bzr200.debian.tar.gz
19:36:14 <pabelanger> ianw: what is needed for that?
19:37:15 <Rocky_g> Capture it in the wayback machine, then kill it
19:37:59 <ianw> robots.txt seems like it allows it
19:39:22 <pabelanger> lets check with fungi when he is back online. But, I tend to agree with mordred, all the data has already been migrated. But, it would be good to understand, why it was still running
19:39:52 <pabelanger> and with that, I'm going to end our meeting.
19:39:55 <mordred> also - jbryce still has a shell account on the machine :)
19:40:10 <pabelanger> #endmeeting