14:00:22 <cdent> #startmeeting placement
14:00:24 <openstack> Meeting started Mon Apr 15 14:00:22 2019 UTC and is due to finish in 60 minutes.  The chair is cdent. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:25 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:28 <openstack> The meeting name has been set to 'placement'
14:00:35 <gibi> o/
14:00:36 <takashin> o/
14:00:41 <tetsuro> o/
14:00:44 <cdent> #link agenda https://wiki.openstack.org/wiki/Meetings/Placement#Agenda_for_next_meeting
14:00:51 <cdent> #topic last meeting
14:00:51 <cdent> #link last minutes http://eavesdrop.openstack.org/meetings/placement/2019/placement.2019-04-08-14.00.html
14:00:57 <edleafe> \o
14:01:10 <cdent> there were no actions from the last meeting, most attention has been on spec reviews and pre-ptg emails
14:01:22 <cdent> anyone have anything from the last meeting they need to revisit?
14:02:04 <alex_xu> o/
14:02:18 <cdent> #topic specs and review
14:02:19 <cdent> #link latest pupdate http://lists.openstack.org/pipermail/openstack-discuss/2019-April/005070.html
14:02:30 <cdent> there was a lot of stuff in the latest pupdate
14:02:40 <cdent> so much that I created a little tool to go back and look at it all easily:
14:02:48 <efried> ol
14:02:54 <cdent> #link placement-reviewme https://github.com/cdent/placement-reviewme
14:03:14 <cdent> which may be totally contrary to other people's workstyles, but I figured it might be useful to someone else so there ya go
14:04:03 <sean-k-mooney> it might be useful to have it jsut generate a gerrity query for all the reviews
14:04:09 <cdent> Of the various specs, the two of gibi's that I've re-proposed needs some attention:
14:04:14 <cdent> #link any trait https://review.openstack.org/#/c/649992/
14:04:15 <cdent> #link mixing required traits https://review.openstack.org/#/c/649368/
14:04:31 <gibi> cdent: I will try to look at it but cannot promise
14:04:33 <cdent> sean-k-mooney: this is what I mean by different work styles. The pupdate services as a human inetermediary/filter
14:04:45 <sean-k-mooney> cdent: :)
14:04:59 <cdent> curation and all that
14:05:20 <sean-k-mooney> yep i just click the links :) anyway thats off topic
14:06:36 <sean-k-mooney> is there a reason those too specs are seperate
14:06:43 <cdent> any other specs that people want to draw attention to (in placement or anywhere else)
14:06:48 <sean-k-mooney> i assume we woudl want to impleement both?
14:06:49 <cdent> sean-k-mooney: because nobody has reviewed them yet
14:07:05 <sean-k-mooney> ah ok
14:07:59 <cdent> #topic bugs
14:07:59 <cdent> #link Placement bugs (launchpad) https://bugs.launchpad.net/nova/+bugs?field.tag=placement
14:08:00 <cdent> #link Placement storyboard https://storyboard.openstack.org/#!/project_group/placement
14:08:16 <cdent> last I looked (this morning) there were no stand out bugs. anyone have any concerns?
14:08:59 <cdent> #topic PTG/Forum
14:09:05 <cdent> #link etherpad for ptg https://etherpad.openstack.org/p/placement-ptg-train
14:09:16 <efried> how are we doing on that, overall, would you say?
14:09:26 <cdent> The pre-ptg emails are going pretty well. some of them have not yet had any responses, but we've made some headway
14:09:44 <efried> Like, how much stuff do we still have that we would wind up discussing at the PTG if we could find time where we weren't all in other rooms?
14:09:45 <cdent> I think it might be useful to sort of digest a few of them
14:10:06 <efried> I feel like you've been doing that to some extent. If not digesting, at least closing.
14:10:11 <cdent> because it feels like there's still quite a bit to discuss, but people aren't actively discussing
14:10:30 <cdent> yeah, the ones that are clearly nearing an end, I've tried to close
14:10:42 <cdent> but there are others which need a bit of a push
14:10:53 <cdent> and some (which I'll link in a moment) that haven't really started
14:11:01 <cdent> #link envisioning other tech http://lists.openstack.org/pipermail/openstack-discuss/2019-April/004789.html
14:11:07 <cdent> #link Use in_tree getting allocation candidates http://lists.openstack.org/pipermail/openstack-discuss/2019-April/004873.html
14:11:11 <cdent> #link Testing PlacementFixture effectively http://lists.openstack.org/pipermail/openstack-discuss/2019-April/004880.html
14:11:15 <cdent> #link Shared resource providers for shared disk on compute hosts http://lists.openstack.org/pipermail/openstack-discuss/2019-April/004881.html
14:11:21 <cdent> #link NUMA Topology with placement http://lists.openstack.org/pipermail/openstack-discuss/2019-April/004884.html
14:12:01 <cdent> What I'll probably do, tomorrow, is go through the whole stack of threads and try to join() them in some sort of fashion to keep them going.
14:12:20 <efried> The not-started ones perhaps aren't cared enough about to warrant pursuing at this time.
14:12:35 <cdent> that's perhaps true
14:13:00 <cdent> the shared disk one I think it's a matter of "the action is elsewhere"
14:13:13 <cdent> probably also true for numa topology
14:13:31 <cdent> although, if reviewing pacing on specs means anything...
14:14:13 <sean-k-mooney> cdent: i think we all want to see both numa and shared disk get resolved. my persecpit of shared disk was no one had time to do it and numa is hard
14:14:43 <cdent> sean-k-mooney: there's a spec under active review:
14:14:48 <sean-k-mooney> numa is hard is also because it need feature not in placment yet if not is done in placemnt rather then with palcement
14:15:01 <cdent> #link shared disk spec: https://review.openstack.org/#/c/650188/
14:15:08 <sean-k-mooney> oh ok
14:15:48 <sean-k-mooney> thanks i was not aware
14:16:10 <cdent> this is why there's a human curated pupdate, so that there can be list of such things ;)
14:17:23 <cdent> gibi: you've expressed some concerns about processing the volume of email. Is there anything we could be doing different to make things a bit cleaner or clearer?
14:18:01 <gibi> cdent: the issue is on my side, I cannot process this amount of information with my limited brain and focus time
14:18:18 <gibi> cdent: I guess it is similar to the ptg shock
14:18:21 <sean-k-mooney> cdent: i have felt similar with the virutal f2f
14:18:34 <gibi> cdent: it is just during the ptg I only focus on the ptg
14:18:45 <gibi> cdent: while now I have many thing in progress
14:18:49 <sean-k-mooney> i lost the lates pupdate in the set of placement emails to the list
14:18:50 <gibi> cdent: still I think the mails are good
14:19:00 <gibi> cdent: and I try to adapt
14:19:31 <cdent> one thing to remember is that you can comment at any point in the process. today, next week, doesn't matter.
14:19:44 <cdent> don't let it rush you
14:19:48 <cdent> it'll keep
14:20:11 <sean-k-mooney> cdent: i did wonder however if gerrit would have been simpler to keep track of for soem of the discussions
14:20:28 <gibi> cdent: ack
14:20:38 <sean-k-mooney> part of that is my email workflow however weher i do not use threading.
14:20:54 <sean-k-mooney> i think the emails have definetly helped
14:21:36 <cdent> the main reason I wanted to stay out of gerrit for most of this stuff is because I wanted to make sure we were able to permeate the boundaries between different things to be a better sense of the big picture, rather than honining on details
14:21:53 <cdent> that's not been a complete success, but I think somewhat useful
14:22:08 <sean-k-mooney> ya it has
14:22:42 <edleafe> Gerrit would be fine for discussing a finished proposal, not for a discussion with many possibilities
14:22:55 <sean-k-mooney> i dissagree with that
14:23:02 <sean-k-mooney> gerrit can be used for both
14:23:20 <sean-k-mooney> but you are right it can be eaier to brainstome without it
14:23:25 <cdent> The thing I think we have to be aware of and conscious of is that we're not legitmately going ot have any significant time in Denver. We have two days for placement. One of them is the last day when we will either be totally drained or people will be gone (jaypipes is leaving friday afternoon if I remember right?)
14:24:01 <cdent> so even if this warmup we are doing now is solely warmup, it is better than nothing
14:24:12 <sean-k-mooney> yes i assume there will be some placement fourme seesions too?
14:24:41 <cdent> just one: about extraction
14:24:58 <edleafe> placement history, not future :)
14:25:18 <cdent> no project update or onboarding sessions were scheduled because of governance
14:25:34 <sean-k-mooney> oh ok i was thinking so large future acrhiterure topics could be covered at fourm and tacticl Train items could be coverd in PTG
14:25:47 <cdent> sean-k-mooney: for nova-related things, yes
14:25:52 <cdent> but outside of that, not really
14:26:09 <cdent> we have some time scheduled with ironic and blazar saturday morning
14:27:05 <cdent> However, this is all okay: given the distribution of members of the team, we're all going to have more obligations outside placement than within it. We just need to make sure that we keep placement up with whatever those outside demands are (bug fixes, etc).
14:27:09 <cdent> If we get to _no_ features this cycle, that's perfectly okay
14:27:22 <cdent> as long as we enable other people to use placement well
14:27:33 <cdent> if we do manage some features, groovy
14:28:19 <sean-k-mooney> yes the hoe to getstart with placment topic on teh mailing list could have made a good fishbool session at the fourm
14:28:19 <cdent> If somebody wants more than that, they can send some people
14:28:57 <sean-k-mooney> kindof like a project unboardign but for potential consumers of placement
14:29:19 <sean-k-mooney> i guess that is what teh ironic/balzar session will be
14:29:28 <cdent> possibly, yes
14:30:08 <cdent> anybody have other things to say about ptg, forum, pre-ptg?
14:30:35 <cdent> oh, I guess i do: I scheduled us for a picture but we can bail on it if people are sick and tired of such things
14:31:35 <cdent> and do we want to try to do a team dinner or anything like that? It feels hard to fit in (given other obligations)
14:31:52 <cdent> I'd be more inclined to have pizza or something like that delivered during the day on saturday
14:32:06 <edleafe> Dinner sounds like a stretch. Photo is fine
14:32:45 <cdent> k
14:32:54 <cdent> #topic opens
14:33:13 <cdent> I'm gone next week. No pupdate this friday or next, won't be here for meeting Monday.
14:33:47 <cdent> I'll try to summarize the pre-ptg emails this thursday and then again the saturday/sunday before the summit. But otherwise I hope to be properly gone.
14:35:10 <gibi> cdent: have a nice time off
14:35:54 <cdent> and a review request: this is cleanups to osc-functional tests to use placement fixture. efried has already expressed that he likes it but feels a bit under-qualified. If you either feel qualified or don't care about being qualified please look at it https://review.openstack.org/#/c/651939/
14:35:56 <sean-k-mooney> cdent: enjoy the easter break
14:35:57 <cdent> thanks gibi
14:36:11 <cdent> I'll be traversing various parts of my family in the midwestern us
14:37:02 <sean-k-mooney> cdent: i feel even less qualifed to review stuff related to the placement fixture and osc but ill try to take a look
14:37:16 <cdent> thanks
14:37:29 <efried> maybe between the two of us sean-k-mooney and I can squeeze out a +2
14:37:29 <cdent> anyone else for opens or anything else?
14:38:01 <efried> (I'm not, however, going to collaborate with sean-k-mooney to squeeze out a #2)
14:38:22 * cdent looks at channel name
14:38:30 <efried> ogod, /me apologizes
14:38:44 <efried> I was planning on taking the day off, clearly that's needed.
14:38:49 <cdent> hehe
14:38:52 <sean-k-mooney> :)
14:39:15 <cdent> I think we better close things up
14:39:20 <cdent> last words?
14:39:57 <cdent> #endmeeting