18:00:33 <krtaylor> #startmeeting third-party 18:00:34 <openstack> Meeting started Mon Oct 27 18:00:33 2014 UTC and is due to finish in 60 minutes. The chair is krtaylor. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:35 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:00:38 <openstack> The meeting name has been set to 'third_party' 18:00:47 <mmedvede> o/ 18:00:50 <sweston> o/ 18:00:55 <asselin> hi 18:01:00 <krtaylor> third-party role call 18:01:26 <VijayTripathi> Hello everyone 18:02:00 <krtaylor> hi all, welcome, let's get started with the meeting 18:02:07 <krtaylor> #topic Welcome & Reminder of OpenStack Mission 18:02:16 <krtaylor> #info The OpenStack Open Source Cloud Mission: to produce the ubiquitous Open Source Cloud Computing platform that will meet the needs of public and private clouds regardless of size, by being simple to implement and massively scalable. 18:02:39 <krtaylor> and here is the agenda for today: 18:02:42 <krtaylor> #link 18:02:45 <krtaylor> oops 18:02:53 <krtaylor> #link https://wiki.openstack.org/wiki/Meetings/ThirdParty#10.2F27.2F14 18:02:57 <krtaylor> that's better 18:03:22 <krtaylor> #topic Review of previous week's open action items 18:03:31 <krtaylor> no action items from last week 18:03:42 <krtaylor> #topic Announcements 18:03:50 <krtaylor> any announcements? 18:04:08 <krtaylor> I don't have any 18:04:32 <krtaylor> although I guess I do, I want to remind everyone that we will not have a meeting next week 18:04:38 <krtaylor> due to summit 18:04:51 <krtaylor> I have put that on our meeting agenda page 18:05:03 <krtaylor> but a reminder - no meeting next week 18:05:31 <krtaylor> #topic OpenStack Program items 18:06:11 <krtaylor> only thing here is the summit session, reminder to get it added to your schedule if you are going, I'd like to get everyone's input 18:06:45 <krtaylor> #topic Deadlines & Deprecations 18:07:03 <krtaylor> none here either 18:07:15 <krtaylor> #topic Highlighting a Program or Gerrit Account 18:07:57 <krtaylor> I added the links for the puppet module split fest as described in last weeks meeting 18:08:11 <krtaylor> asselin, how are you coming with your split? 18:08:58 <asselin> I got it started, and updated the spec with more clear instructions: https://review.openstack.org/#/c/129768/ 18:09:12 <krtaylor> if you want, you can set the topic to module-split so we can track it all together, but not abig deal 18:09:15 <krtaylor> ok, great 18:09:33 <asselin> I'm going to go through the rest of the stepps today & ref some of the patches I see in today's meeting agenda. 18:09:43 <asselin> sure, will try that 18:10:08 <asselin> topic updated 18:10:10 <krtaylor> asselin, mmedvede has done a similar steps list for our team internally 18:10:19 <anteaya> asselin: let me know when you have something and we can time the freeze for that module 18:10:24 <krtaylor> but I see he has commented on yours 18:11:08 <asselin> anteaya, thanks. I'd like to go through all the steps, then time the freeze, and then redo the 1st steps to sync up the private repo with the frozen one. 18:11:59 <asselin> since those first few steps are easy to do. 18:12:00 <mmedvede> asselin: with my split (kibana) I did not have to do any sort of freeze, but that is because it is slow-moving repository 18:12:29 <anteaya> asselin: good plan 18:12:32 <asselin> mmedvede, good point. It's possible there were no changes either to mine (jenkins) 18:12:45 <krtaylor> also mmedvede has noted that there is potentially some loss of git log with git subtree split 18:12:57 <anteaya> kibana is lower activity then jenkins 18:13:05 <anteaya> mmedvede: has your split merged? 18:13:36 <mmedvede> anteaya: yes, first one have merged. The second one (pulling out the module from system-config) is pending 18:13:41 <krtaylor> #link https://review.openstack.org/#/q/branch:master+topic:module-split,n,z 18:13:47 <krtaylor> ^^^ shows everything 18:14:24 <anteaya> mmedvede: good work! 18:14:39 <mmedvede> anteaya: jenkins was failing on the second patch (we discussed it in infra), I have fixed it, it seems 18:14:39 <krtaylor> ++ 18:14:43 <mmedvede> anteaya: thanks 18:14:56 <krtaylor> I'm not done with my pair yet, second one is pending 18:15:12 <anteaya> mmedvede: I'm only around intermittently today, but if the second patch is ready to go, get some attention on it 18:15:28 <anteaya> krtaylor: good 18:15:33 <mmedvede> anteaya: ok 18:16:02 <mmedvede> it is ready to go, so I would ask in -infra later 18:16:16 <anteaya> mmedvede: yes 18:16:33 <krtaylor> yea! I see it passed jenkins 18:16:50 <krtaylor> mmedvede, did you want to comment on the log loss? 18:17:21 <mmedvede> krtaylor: sure 18:17:56 <mmedvede> FYI, the 'git subtree' split method will lose history in case there were renames 18:18:07 <mmedvede> Something to be aware of. 18:18:47 <mmedvede> There is a way around it, and I can describe it if needs to be. But it takes time. And only worth if there is a lot of commits that are left our by subtree 18:19:40 <mmedvede> krtaylor: that is all I wanted to mention 18:19:41 <krtaylor> mmedvede, I guess we can wait and see, maybe add to the spec for instructions to fix 18:19:51 <krtaylor> mmedvede, thanks 18:20:17 <krtaylor> anything else on the module split? 18:20:31 <krtaylor> ok, onward then 18:21:11 <krtaylor> fungi asked to mention the removal of pypi mirror from static.openstack.org 18:21:26 <krtaylor> #link https://review.openstack.org/#/c/119466/ 18:22:16 <fungi> yes, there are apparently third-party ci operators who are, knowingly or not, using our pypi mirrors instead of running their own or using pypi.python.org 18:22:50 <krtaylor> ah, ok, thanks fungi 18:23:06 <fungi> the citrix xenserver ci broke and had b=to be disabled today because it was still using pypi.openstack.org which is no longer updated since quite a while now 18:23:51 <fungi> anyway, that's all. just wanted to make sure everyone was aware to keep an eye out for that in your configurations 18:24:06 <krtaylor> I saw the disable, that explains it 18:24:15 <krtaylor> thanks fungi 18:25:10 <krtaylor> any questions on this? 18:25:27 <krtaylor> ok, then next 18:25:33 <krtaylor> #topic Open Discussion 18:26:22 <krtaylor> ok, not everyone at once :) 18:26:32 <anteaya> as an fyi I am not going to be availabe the Monday after summit 18:26:48 <anteaya> not that you need me, you are doing fine on your own 18:26:49 <fungi> oh, naggappan was asking for third-party ci configuration help in #openstack-dev earlier today. i suggested showing up for this, but looks like the recommendation was ignored 18:26:53 <anteaya> good work krtaylor 18:27:04 <anteaya> fungi: thanks for trying 18:27:07 <krtaylor> OH, I forgot to add CI self check to the agenda 18:27:10 <krtaylor> anteaya, thanks 18:27:27 <asselin> I suppose no meeting next monday. 18:27:43 <fungi> also, seems the openstack meetbot isn't updating the channel topic 18:27:47 <anteaya> I think we decided no meeting next monday 18:27:55 <anteaya> but haven't announced it anywhere 18:27:58 <krtaylor> asselin, yes I added that to the agenda page as a reminder also 18:28:09 <anteaya> fungi: good eye, I hadn't noticed 18:28:14 <dougwig> to double-check, devstack-gate isn't using a bad pypi mirror, right? 18:28:18 <anteaya> fungi: I wonder why 18:28:33 <anteaya> fungi: the meetbot commands seem accurate 18:29:06 <krtaylor> hm, yeah 18:29:13 <krtaylor> #topic test 123 18:29:17 <krtaylor> nope 18:29:18 <fungi> anteaya: it's missing channel operator perms. probably lost them in a netslipt. i'm fixing now 18:29:31 <anteaya> fungi: thanks 18:29:41 <clarkb> dougwig: not in the upstream testing. copy_mirror_config in devstack-gate expects a working mirror setup that can be copied though 18:29:55 <clarkb> dougwig: our nodepool does this for us but puppet or any other config management could too 18:30:00 <krtaylor> #topic Open Discussion 18:30:07 <krtaylor> yea, that worked 18:30:15 <krtaylor> thanks fungi 18:30:19 <fungi> sure thing 18:31:57 <krtaylor> let's see, last week I noted that hyper-v had a static IP, but I haven't checked to see if that got fixed 18:33:11 <krtaylor> did anyone else notice any CI systems with strange results comments, ect? 18:34:22 <krtaylor> hyper-v did reply to my email that they are going to fix their IP 18:34:36 <krtaylor> ok, everyone is quiet today 18:35:02 <krtaylor> if we are winding down then I'll let everybody have 25 minutes back 18:35:09 <krtaylor> anything else? 18:35:31 * krtaylor likes the calm week before summit 18:35:33 <asselin> my devstack-gate changes are still pending reviews 18:35:45 <krtaylor> asselin, link? 18:36:09 <asselin> https://review.openstack.org/#/c/123578/ 18:36:20 <asselin> https://review.openstack.org/#/c/122896/ 18:36:43 <asselin> I've been cherry-picking them for my own ci system 18:37:05 <asselin> maybe I should change the topic? 18:37:12 <anteaya> asselin: the second one is ready to go 18:37:19 <anteaya> find sdague and request a +A 18:37:23 <mmedvede> asselin: I would like to have that merged 18:37:34 <krtaylor> ++ 18:37:37 <asselin> need another core no? 18:38:04 <anteaya> asselin: yes and sdague is a core 18:38:05 <asselin> (after it merges, I'll rebase the dependent one....) 18:38:29 <anteaya> oh the parent is ready to go 18:38:34 <anteaya> isn't 18:38:35 <asselin> https://review.openstack.org/#/c/123578/ needs to merge first and only has 1 core 18:39:10 <asselin> which is sdague 18:39:11 <krtaylor> both have one core 18:39:33 <ociuhandu> krtaylor: sorry, missed the first part due to the hour change here :( We sent out the request for the FQDN in Microsoft, hope to have that sorted out asap 18:39:48 <anteaya> asselin: yes, sorry I mis-understood 18:39:55 <asselin> anteaya, np 18:40:19 <asselin> need to find more devstack-gate cores 18:40:40 <krtaylor> ociuhandu, no worries, yes I mentioned that you had replied 18:41:02 <anteaya> asselin: it is sdague and infra-core 18:41:07 <asselin> jhesketh can you look at https://review.openstack.org/#/c/123578/1 18:41:11 <krtaylor> hm, maybe thats why it is so light today, time change 18:41:12 <asselin> anteaya, ok thanks 18:41:15 <anteaya> asselin: he is asleep 18:41:42 <anteaya> jhesketh is in tasmania 18:41:47 <asselin> ok I can ping -infra after the meeting 18:41:55 <anteaya> asselin: kk 18:42:04 <krtaylor> sounds like a plan 18:42:12 <clarkb> note I already "reviewed" those changes and dn't see a reasonfor them to exist 18:42:21 <clarkb> the first is a reasonable refactor but the second is unnecessary 18:42:41 * krtaylor looks 18:44:12 <krtaylor> asselin, any questions for clarkb 18:44:51 <asselin> clarkb and I discussed it a while ago. we agreed to get more opinions for other cores, if i remember right. 18:44:56 <clarkb> yup 18:44:59 <asselin> clarkb, maybe you could add you comments to the patch? 18:45:06 <clarkb> I can leave them there 18:45:59 <asselin> so for everyone else, the pre-clean hook can just be done as a post-clean hook, which can be done external to the devstack-gate script, although with a bit more complexity 18:46:18 <clarkb> I think its simpler 18:46:20 <clarkb> not more complex 18:46:36 <krtaylor> mmedvede, ^^^ 18:47:34 <asselin> right, we disagree on simple vs. complex, and decided to get more opinions from other cores. 18:47:42 <asselin> (honestly I can go either way....) 18:48:15 <asselin> but still think hooks are simpler :) 18:48:51 <clarkb> the counter to that is we have tons of confusion around the existing hooks. the existing hooks need to run in the context of d-g so we can't really get rid of them 18:49:39 <asselin> anyone else with an opinion? 18:49:44 <clarkb> but there is always a lot of confusion when people add jobs around what they are getting when using a hook. (are all services running? what permissions does current user have, if I override a hook like gate_hook what ends up needing to be done redundanty, and so on) 18:50:07 <clarkb> if however you just run a script prior to and after running d-g the person running that script is in complete control 18:50:11 <clarkb> this is simple and not confusing 18:50:47 <anteaya> clarkb: has had a lot of experience maintaining this sort of thing 18:51:00 <anteaya> especially picking up devs who have fallen down on it 18:51:05 <anteaya> I defer to his opinion 18:51:16 <asselin> ok, so if noone else objects, I can abandon https://review.openstack.org/#/c/122896/ 18:51:40 <dougwig> i'm a case study in not using the hooks correctly, since i use sed to insert my own in-between devstack setup and tempest in d-g. 18:53:04 <asselin> ok, no objections, so no hooks. 18:53:11 <krtaylor> 7 minutes 18:53:13 <asselin> no new hooks 18:53:18 <krtaylor> any other comments? 18:54:13 <krtaylor> ok, well I think we are done here then 18:54:33 <mmedvede> I missed it, I still kind of liked the hooks, but oh well 18:54:36 <krtaylor> remember no meeting next week, and I'll send email reminder 18:54:58 <krtaylor> mmedvede, lets move it to infra 18:55:21 <krtaylor> thanks everyone, hope to see you in Paris! 18:55:27 <sweston> thanks, everyone !! 18:55:30 <dougwig> bye 18:55:41 <krtaylor> #endmeeting