09:00:23 <priteau> #startmeeting blazar 09:00:24 <openstack> Meeting started Tue Sep 25 09:00:23 2018 UTC and is due to finish in 60 minutes. The chair is priteau. Information about MeetBot at http://wiki.debian.org/MeetBot. 09:00:25 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 09:00:27 <openstack> The meeting name has been set to 'blazar' 09:01:01 <priteau> #topic RollCall 09:01:52 <bertys_> o/ 09:02:08 <priteau> Hello bertys_ 09:02:16 <priteau> Masahito is on travel this week 09:02:33 <priteau> I don't know about Tetsuro but he is not online. 09:02:34 <bertys_> Right 09:02:40 <priteau> Looks like it is just the two of us. 09:03:25 <priteau> I don't have much on the agenda for today 09:03:27 <priteau> 1. Forum Topic Submission 09:03:31 <priteau> 2. AOB 09:04:22 <priteau> #topic Forum Topic Submission 09:05:22 <priteau> I have started to put our topics in the form needed for the submission 09:05:25 <priteau> #link https://etherpad.openstack.org/p/Berlin-stein-forum-blazar-brainstorming 09:05:33 <priteau> I am not quite done yet 09:05:53 <bertys_> It seems the deadline is tomorrow 09:06:26 <priteau> Yes indeed. 09:07:00 <bertys_> Do you know if john (or others) plan to submit topic related to preemptible instances? 09:08:18 <bertys_> I have not seen anything yet at https://www.openstack.org/summit/berlin-2018/vote-for-speakers/#/22782 09:08:20 <priteau> I asked John about it, he didn't think a Forum session was useful. At the moment they working on updating their specs, most of the discussion will happen on Gerrit 09:08:41 <bertys_> ok got it 09:09:50 <priteau> I see there are two formats possible, fishbowl and collab discussion 09:10:09 <priteau> Is collab discussion like the rooms at the PTG? 09:12:01 <priteau> I am thinking collab discussion might be best for us. 09:12:10 * tetsuro apologize to be late 09:12:51 <priteau> Hi tetsuro! 09:13:20 <tetsuro> priteau: Hi, sorry to be late 09:13:24 <priteau> No problem 09:13:29 <priteau> We're discussing the Forum 09:13:48 <priteau> I've finished writing the text for the Python bindings for the placement API topic 09:14:00 <priteau> Could you please review it? 09:14:05 <priteau> Bottom part of https://etherpad.openstack.org/p/Berlin-stein-forum-blazar-brainstorming 09:15:19 <tetsuro> Yes, I agree that it is a good idea to discuss it in the Forum 09:16:34 <priteau> We need a moderator, would you be willing to do it, since you are our placement expert? 09:17:04 <tetsuro> Sure I can be a moderator for this session. 09:17:08 <priteau> Thanks! 09:20:03 <priteau> I've finished a draft of the other session as well and proposed myself as moderator 09:20:23 <priteau> Please review and make changes if needed 09:21:44 <priteau> I integrated the topics proposed by csatari into the first submission called "Extending reservation to other resource types" 09:22:56 <priteau> Any comment? 09:25:10 <priteau> Masahito said he was at the ONS summit so he's in Amsterdam 09:25:35 <priteau> I will email him asking to review the topics today, hopefully he will have time to answer before tomorrow night 09:26:01 <tetsuro> I'm interested in reserving NUMA-ish things, but have no idea how to realize it. So, it would be nice to get some ideas there. 09:27:28 <priteau> Indeed it would be useful to have a Nova expert on NUMA in the room 09:29:21 <priteau> Thanks for adding a comment about NUMA tetsuro 09:33:23 <priteau> tetsuro: Would you like your comment to be integrated in the Forum submission? I would have to reword it a bit. 09:35:08 <tetsuro> To think it twice, blazar doesn't have to be aware of cpu pinning, so you don't need to add my comments to the submission. sorry. 09:35:41 <tetsuro> But I agree supporting numa is important, I'll rewrite the comment. 09:36:49 <priteau> Thanks 09:38:53 <priteau> Are NUMA nodes represented in placement already? 09:40:15 <tetsuro> Placement is ready to model NUMA, but who puts NUMA info is Nova. Nova is not yet ready to report NUMA-ish things to placement. 09:40:47 <priteau> OK 09:41:14 <tetsuro> So the problem is how blazar takes that NUMA info into blazar's DB. 09:41:26 <priteau> I found this related spec: https://review.openstack.org/#/c/555081/ 09:42:47 <tetsuro> Yup this is the spec for modeling pinning CPUs in placement. 09:43:26 <tetsuro> But it's not high priority in Nova in Stein. 09:43:35 <priteau> Any other comment before moving to AOB? 09:43:43 <tetsuro> Nope. 09:43:55 <priteau> #topic AOB 09:44:46 <priteau> Anyone has an update to share? bertys_ I saw you were working on the Ansible module 09:46:08 <bertys_> priteau: For the Blazar role, some work is ongoing related to rabbitmq and backported to rocky 09:47:13 <priteau> That's great news 09:48:31 <priteau> I only looked quickly, will review more closely lalter 09:49:30 <priteau> On my side I've been working on getting University of Chicago folks involved in review and submitting patches, hopefully we will get more activity from them. 09:50:40 <priteau> tetsuro do you have anything to share? 09:51:03 <tetsuro> Nope. 09:51:24 <priteau> No problem :-) 09:51:38 <priteau> If that's all should we finish the meeting early? 09:52:45 <tetsuro> http://lists.openstack.org/pipermail/openstack-dev/2018-September/134848.html 09:53:31 <priteau> Right I saw that, congrats! 09:53:55 <tetsuro> chris has nominated me for placement-core team. It's not decided, but for your information. 09:54:09 <tetsuro> thanks, priteau :) 09:55:32 <priteau> Thanks for joining everyone 09:55:36 <priteau> #endmeeting