*** tetsuro has joined #openstack-placement | 01:01 | |
*** cdent has quit IRC | 01:01 | |
mriedem_away | done | 01:12 |
---|---|---|
*** mriedem_away has quit IRC | 01:15 | |
*** tetsuro has quit IRC | 01:20 | |
openstackgerrit | Merged openstack/placement master: Upgrade from rocky-nova docs https://review.openstack.org/644396 | 02:45 |
openstackgerrit | Merged openstack/placement master: Add prelude to release notes https://review.openstack.org/644567 | 02:45 |
openstackgerrit | Merged openstack/placement master: Link to more info on service user and endpoints from deployment https://review.openstack.org/644568 | 02:45 |
*** e0ne has joined #openstack-placement | 07:17 | |
*** e0ne has quit IRC | 07:27 | |
*** e0ne has joined #openstack-placement | 07:42 | |
*** alex_xu has quit IRC | 08:04 | |
*** ttsiouts has joined #openstack-placement | 08:05 | |
*** alex_xu has joined #openstack-placement | 08:06 | |
*** e0ne has quit IRC | 08:15 | |
*** helenafm has joined #openstack-placement | 08:19 | |
*** ttsiouts has quit IRC | 08:26 | |
*** ttsiouts has joined #openstack-placement | 08:34 | |
*** ttsiouts has quit IRC | 08:52 | |
*** ttsiouts has joined #openstack-placement | 08:54 | |
openstackgerrit | Balazs Gibizer proposed openstack/nova-specs master: Update Network Bandwidth resource provider spec https://review.openstack.org/644810 | 10:14 |
*** cdent has joined #openstack-placement | 10:27 | |
cdent | stephenfin, gibi : If you get. chance to look at https://review.openstack.org/#/c/644690/ and https://review.openstack.org/#/c/644687/ this morning that would be grand | 10:39 |
gibi | cdent: looking | 10:39 |
gibi | cdent: left feedback | 10:45 |
cdent | thanks | 10:45 |
*** e0ne has joined #openstack-placement | 10:45 | |
*** ttsiouts has quit IRC | 11:21 | |
*** ttsiouts has joined #openstack-placement | 11:37 | |
openstackgerrit | Chris Dent proposed openstack/placement master: Address followups in the upgrade from nova doc https://review.openstack.org/644690 | 11:42 |
*** ttsiouts has quit IRC | 11:42 | |
cdent | stephenfin: grammar bits tuned ^ | 11:42 |
stephenfin | cdent: +2 | 11:43 |
cdent | stephenfin: feel free to proxy eric's +2 to give it the +W | 11:43 |
cdent | as I'd like to cut the release today if possible | 11:43 |
stephenfin | Good point. Done and done | 11:44 |
cdent | thanks | 11:44 |
cdent | if you can provide some comments on https://review.openstack.org/#/c/644687/ I'll incorporate them and gibi's into a new version before we are joined by mriedem | 11:45 |
*** tssurya has joined #openstack-placement | 11:47 | |
stephenfin | I can but I won't have a chance for a while. Meetings | 11:50 |
cdent | cool, thanks | 11:53 |
openstackgerrit | Merged openstack/os-resource-classes master: Update master for stable/stein https://review.openstack.org/644030 | 12:02 |
*** ttsiouts has joined #openstack-placement | 12:04 | |
openstackgerrit | Merged openstack/os-traits master: Update master for stable/stein https://review.openstack.org/644033 | 12:05 |
*** e0ne has quit IRC | 12:07 | |
openstackgerrit | Merged openstack/os-resource-classes stable/stein: Update .gitreview for stable/stein https://review.openstack.org/644028 | 12:29 |
openstackgerrit | Merged openstack/os-traits stable/stein: Update .gitreview for stable/stein https://review.openstack.org/644031 | 12:30 |
openstackgerrit | Merged openstack/os-resource-classes stable/stein: Update UPPER_CONSTRAINTS_FILE for stable/stein https://review.openstack.org/644029 | 12:30 |
openstackgerrit | Merged openstack/osc-placement master: Update master for stable/stein https://review.openstack.org/644045 | 12:34 |
openstackgerrit | Merged openstack/os-traits stable/stein: Update UPPER_CONSTRAINTS_FILE for stable/stein https://review.openstack.org/644032 | 12:36 |
openstackgerrit | Merged openstack/osc-placement stable/stein: Update .gitreview for stable/stein https://review.openstack.org/644043 | 13:02 |
openstackgerrit | Merged openstack/osc-placement stable/stein: Update UPPER_CONSTRAINTS_FILE for stable/stein https://review.openstack.org/644044 | 13:03 |
*** e0ne has joined #openstack-placement | 13:16 | |
*** mriedem has joined #openstack-placement | 13:17 | |
mriedem | . | 13:21 |
openstackgerrit | Merged openstack/placement master: Address followups in the upgrade from nova doc https://review.openstack.org/644690 | 13:21 |
cdent | mriedem: I had planned to make some changes to your https://review.openstack.org/#/c/644687/ but then didn't get to it because a) other stuff b) stephenfin said he was going to give it a review so figured better to wait. so it is still pending | 13:44 |
cdent | other than that we seem in good shape on placement side | 13:44 |
*** helenafm has quit IRC | 13:48 | |
mriedem | cdent: gibi: replied in that one, i'm just going to remove the 'list resource providers' part since it's more confusing than helpful | 13:49 |
cdent | wfm | 13:49 |
mriedem | listing resource classes and traits is good enough | 13:49 |
cdent | does prove it worked | 13:49 |
gibi | mriedem: works for me | 13:50 |
gibi | mriedem: I assume traits and RCs are populated automatically from os-* libs so those lists are not empty even if no nova compute is running | 13:51 |
openstackgerrit | Matt Riedemann proposed openstack/placement master: Flesh out the post-install verify doc https://review.openstack.org/644687 | 13:51 |
mriedem | correct | 13:51 |
cdent | mriedem: is there policy associated with who is supposed to be/allowed on the stable-maint group? | 14:04 |
cdent | I discovered placement-stable-maint wasn't set up, so got that going, added me, and the stable-maint-core list. if/when we wanted to make it more, are there rules? | 14:04 |
mriedem | cdent: yeah https://docs.openstack.org/project-team-guide/stable-branches.html#stable-maintenance-teams | 14:06 |
cdent | ah, cool, thanks | 14:06 |
mriedem | although those are becoming more lenient over time | 14:06 |
mriedem | tl;dr the project stable team proposes new cores to the stable-maint-core team | 14:06 |
mriedem | since stable core is not the same as core on master - there are different rules | 14:07 |
mriedem | but that was also when stable-maint-core was much more actively involved in lots of project's stable branches, which they are less now, so it's just becoming more of a trust model | 14:07 |
mriedem | goes back to the age old "i'm ptl on project x but i'm not a stable core on my own project, why not?" thing | 14:07 |
cdent | well in this case, I pulled the trigger because somebody needed to merge all the bot stuff on os-resource-classse, os-traits, and osc-placement and I discovered I couldn't | 14:08 |
cdent | otherwise I'd happily let it slide | 14:08 |
*** ttsiouts has quit IRC | 14:10 | |
*** ttsiouts has joined #openstack-placement | 14:10 | |
*** ttsiouts has quit IRC | 14:15 | |
*** ttsiouts has joined #openstack-placement | 14:19 | |
mriedem | i'm on stable-maint-core so feel free to ping me for things | 14:33 |
cdent | thanks mriedem. i'm sure I will. At this point: nothing to do \o/ | 14:48 |
*** altlogbot_3 has joined #openstack-placement | 15:12 | |
*** edleafe_ has joined #openstack-placement | 15:14 | |
openstackgerrit | Dan Smith proposed openstack/nova-specs master: Add request-filter-image-types spec https://review.openstack.org/644625 | 15:42 |
mriedem | stephenfin: question in https://review.openstack.org/#/c/641422/ | 15:46 |
stephenfin | mriedem: Yeah, exactly what you said | 15:47 |
stephenfin | It's noise, but I figured there was some value for direct linking. I'd be happy to see it go too though | 15:47 |
stephenfin | (but that patch was already big enough) | 15:48 |
mriedem | ack | 15:55 |
mriedem | at this point you might want to split your change and rebase on master | 15:55 |
mriedem | s/split/decouple/ | 15:55 |
cdent | yeah, I said that in the parent too | 16:01 |
efried | cdent: I didn't touch the placement etherpad, but I suspect some of those topics can be moved to the nova/placement xproj | 16:03 |
cdent | likely yeah | 16:03 |
*** efried is now known as efried_rolling | 16:03 | |
cdent | stephenfin: if you're going to rebase that patch, shout out in here, and I'll wait on it (there are two other patches still crawling through the gate) | 16:10 |
openstackgerrit | Merged openstack/placement master: Centralize registration of logging options https://review.openstack.org/644591 | 16:13 |
openstackgerrit | Merged openstack/placement master: Flesh out the post-install verify doc https://review.openstack.org/644687 | 16:13 |
stephenfin | cdent: which patch is this? | 16:20 |
stephenfin | oh, nvm | 16:20 |
openstackgerrit | Dan Smith proposed openstack/nova-specs master: Add request-filter-image-types spec https://review.openstack.org/644625 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/placement master: Group API versions by release https://review.openstack.org/641422 | 16:30 |
stephenfin | cdent, mriedem: as requested | 16:30 |
stephenfin | ^ | 16:30 |
mriedem | stephenfin: comment inline, | 16:33 |
mriedem | but i'm not sure why you're, in the same change, making unrelated format changes to the section titles themselves | 16:33 |
mriedem | i.e. 'member_of' | 16:33 |
mriedem | is that just a rebase snafu? | 16:33 |
stephenfin | that is, yup | 16:36 |
* stephenfin goes to fix again | 16:37 | |
stephenfin | mriedem: Actually, I'm not sure if there's much value in it. I already have to change some headers (to change the level, remove the 'Maximum in FOO' parts) so the single quotes probably aren't much extra | 16:38 |
stephenfin | *much value in not undoing those particular changes | 16:38 |
*** altlogbot_3 has joined #openstack-placement | 16:40 | |
cdent | stephenfin: it looks like it is inconsistent: some of it is ' some ``. I don't reckon it is a huge deal one way or another, but if you're fixing... | 16:41 |
mriedem | welp i'm going to have to compare the before and after now | 16:41 |
*** altlogbot_3 has quit IRC | 16:41 | |
openstackgerrit | Stephen Finucane proposed openstack/placement master: Group API versions by release https://review.openstack.org/641422 | 16:43 |
stephenfin | cdent: Oh, weird. Fixed | 16:43 |
stephenfin | I don't know why they didn't show up as merge conflicts. Presumably because I'd made additional changes to efried | 16:43 |
cdent | sometimes after so many rebases magical things happen | 16:45 |
*** N3l1x has joined #openstack-placement | 17:00 | |
*** e0ne has quit IRC | 17:07 | |
*** altlogbot_3 has joined #openstack-placement | 17:26 | |
*** altlogbot_3 has quit IRC | 17:28 | |
*** altlogbot_3 has joined #openstack-placement | 17:30 | |
*** ttsiouts has quit IRC | 17:40 | |
*** ttsiouts has joined #openstack-placement | 17:41 | |
*** ttsiouts has quit IRC | 17:46 | |
*** e0ne has joined #openstack-placement | 18:17 | |
*** irclogbot_2 has quit IRC | 18:21 | |
*** irclogbot_0 has joined #openstack-placement | 18:24 | |
*** irclogbot_0 has quit IRC | 18:27 | |
*** irclogbot_1 has joined #openstack-placement | 18:29 | |
openstackgerrit | Merged openstack/placement master: Group API versions by release https://review.openstack.org/641422 | 18:30 |
*** e0ne has quit IRC | 18:53 | |
*** altlogbot_3 has quit IRC | 19:00 | |
*** altlogbot_1 has joined #openstack-placement | 19:02 | |
*** altlogbot_1 has quit IRC | 19:06 | |
*** tssurya has quit IRC | 19:07 | |
*** altlogbot_2 has joined #openstack-placement | 19:07 | |
*** efried_rolling is now known as efried | 19:11 | |
*** irclogbot_1 has quit IRC | 19:16 | |
*** tssurya has joined #openstack-placement | 19:16 | |
*** irclogbot_2 has joined #openstack-placement | 19:17 | |
*** irclogbot_2 has quit IRC | 19:30 | |
*** irclogbot_1 has joined #openstack-placement | 19:31 | |
*** altlogbot_2 has quit IRC | 19:34 | |
*** altlogbot_3 has joined #openstack-placement | 19:36 | |
sean-k-mooney | cdent: actully semi plament question that i would like your input on if you have a sec | 19:57 |
cdent | sure | 19:57 |
sean-k-mooney | so nova supports specifying the virtual video device model we use as the graphic adapater in the libvirt dirver | 19:58 |
sean-k-mooney | https://libvirt.org/formatdomain.html#elementsVideo | 19:58 |
sean-k-mooney | and we have an enuma of the allowed vlaues | 19:58 |
sean-k-mooney | im planning to addd virtio to that enum this cycle | 19:58 |
sean-k-mooney | e.g. in train | 19:58 |
sean-k-mooney | shoudl the supported models be repored/tracked in placment in some way | 19:59 |
* cdent thinks | 19:59 | |
sean-k-mooney | e.g. as traits againt the compute node resoure provider | 19:59 |
sean-k-mooney | there is basicaly the same question for vif_model to | 20:00 |
cdent | assuming it is only a quality, without any indicatin of quantity, then trait would make sense because you're saying "this compute node can do the following video device types" | 20:00 |
cdent | (is that what you're saying?) | 20:01 |
sean-k-mooney | yes i am saing this compute cand support virtio and qxl because it has libvirt greate then 1.3.0 | 20:01 |
sean-k-mooney | but ut cant do gop or none because they werent added until later and this node does not have a new enough libvirt | 20:02 |
cdent | yeah, sounds like traits | 20:02 |
cdent | presumably we'd want a namespace in os-traits for such things | 20:02 |
sean-k-mooney | ok that was going to be my next question | 20:02 |
sean-k-mooney | it would be ok to add namespased standard traits for this | 20:02 |
sean-k-mooney | somethign like comptue_libvirt_video_model_virtio | 20:03 |
sean-k-mooney | or whatever it woudl be | 20:03 |
cdent | it sounds in the same vein as the supported images types that dansmith wants to get working | 20:03 |
sean-k-mooney | yes that is waht made me think of it | 20:03 |
cdent | I can't begin to guess at this time what the right namespace would be, but something... | 20:03 |
sean-k-mooney | i was orginally just going to add the virtio value to the enum but then i wondered if i should expose it as a trait or not | 20:04 |
sean-k-mooney | interface model | 20:04 |
sean-k-mooney | wrong terminal | 20:04 |
cdent | edleafe and jaypipes usually have pretty good ideas on namespaces | 20:05 |
cdent | I fairly reliably do _not_ | 20:05 |
sean-k-mooney | im basically extending this image metadta value to allow virtio https://github.com/openstack/glance/blob/master/etc/metadefs/compute-libvirt-image.json#L45-L55 | 20:05 |
* cdent nods | 20:06 | |
sean-k-mooney | we also have the same thing for virtualised nics below it | 20:06 |
sean-k-mooney | https://github.com/openstack/glance/blob/master/etc/metadefs/compute-libvirt-image.json#L68-L82 | 20:06 |
cdent | presumably we've got trait namespaces for that stuff already | 20:06 |
sean-k-mooney | i dont think we do but we could eaily basically copy past it | 20:06 |
sean-k-mooney | anyway it sound like this is not insane and i should praobly write something up to track this | 20:07 |
sean-k-mooney | was there any desion made on how to track os-trait things? storyborad? | 20:07 |
edleafe | I think that if you look at the current namespaces in os-traits, you'll get a good idea where these should fit in | 20:08 |
sean-k-mooney | edleafe: ya its pretty intuitve | 20:08 |
sean-k-mooney | it shoudl be under the compute namespace somewhere | 20:08 |
edleafe | Well, as long as you understand the underlying thing. | 20:09 |
sean-k-mooney | the vif models woudl goin in here https://github.com/openstack/os-traits/blob/master/os_traits/compute/net.py | 20:09 |
sean-k-mooney | and the video device model with go in a sibling video.py or graphics.py file | 20:10 |
cdent | sean-k-mooney: yeah, storyboard is a good place to start | 20:10 |
cdent | make a story with as much detail as you got for now | 20:10 |
sean-k-mooney | edleafe: jaypipes implemented it but i was the one that suggested using directies to model namesapse so i remember how this works :) | 20:11 |
sean-k-mooney | cdent: ok will do. once i figure out how to use storyboard | 20:11 |
cdent | it's...an experience | 20:11 |
cdent | i'm growing to like it | 20:12 |
sean-k-mooney | ill start by grabing hte nova specs tepmleate and fillig out the reslevet section in an etherpad then copy and paste the relvent bits into storyboard | 20:13 |
sean-k-mooney | usually i find if i can answer all the question in the spec template i can format what is need in any tool i use | 20:14 |
cdent | sean-k-mooney: for the changes in os-traits itself I dont' think we'll need the detail of a spec | 20:14 |
cdent | if you think the changes in nova will require a spec, then that would be the place to do a spec | 20:15 |
sean-k-mooney | mainly becase the spec template makes you think about how you expalin the feature to others and not just what your going to do | 20:15 |
sean-k-mooney | oh i know | 20:15 |
sean-k-mooney | i just want ot think about how this would be used end to end before i start | 20:15 |
* cdent nods | 20:16 | |
sean-k-mooney | the os-traits change are simple as long as people agree these qualify as trait | 20:16 |
sean-k-mooney | that was the main thing i want to sanity check | 20:16 |
cdent | definitely sounds like it | 20:16 |
cdent | they are things you want to know but can't consume | 20:16 |
sean-k-mooney | yes exactly | 20:17 |
sean-k-mooney | and it will remove 2 more case of "no valid host found" | 20:17 |
cdent | woot | 20:18 |
sean-k-mooney | we currently do check to see if they image properties ar valid for the selected host way down in th virt driver during spawn | 20:19 |
cdent | bit late | 20:19 |
efried | cdent: https://review.openstack.org/644992 | 20:20 |
sean-k-mooney | just a bit and the code is pretty dumb too | 20:20 |
sean-k-mooney | https://github.com/openstack/nova/blob/master/nova/virt/libvirt/vif.py#L67-L100 | 20:20 |
cdent | efried: I guess that means that wiki page will need some stuff. I'll get on that...probably tomorrow | 20:22 |
efried | cdent: I can do it if you like. up to you. | 20:23 |
cdent | i got it, no urgent rush | 20:23 |
efried | but technically we don't need to do anything until that patch merges :) | 20:23 |
efried | I'm going to send a note to the ML with the gist of the proposal | 20:23 |
sean-k-mooney | night folks o/ | 20:23 |
cdent | bleeding will be 7th day for some, mriedem | 20:25 |
cdent | efried: coo | 20:25 |
mriedem | cdent: oh that's right | 20:28 |
mriedem | TC suckers | 20:28 |
cdent | it's insane | 20:28 |
cdent | i'm already ready to kill folk after a day | 20:28 |
cdent | by saturday ... I don't even want to think about it | 20:29 |
cdent | efried: went ahead and stubbed out the content for the new wiki page and updated the old one | 20:38 |
*** irclogbot_1 has quit IRC | 20:45 | |
*** irclogbot_3 has joined #openstack-placement | 20:47 | |
openstackgerrit | OpenStack Release Bot proposed openstack/placement stable/stein: Update .gitreview for stable/stein https://review.openstack.org/644999 | 20:47 |
openstackgerrit | OpenStack Release Bot proposed openstack/placement stable/stein: Update UPPER_CONSTRAINTS_FILE for stable/stein https://review.openstack.org/645000 | 20:47 |
openstackgerrit | OpenStack Release Bot proposed openstack/placement master: Update master for stable/stein https://review.openstack.org/645001 | 20:47 |
* cdent feels suddenly od | 20:50 | |
cdent | old | 20:50 |
*** e0ne has joined #openstack-placement | 20:52 | |
*** altlogbot_3 has quit IRC | 20:53 | |
*** altlogbot_2 has joined #openstack-placement | 20:55 | |
efried | There's nothing sudden about your odness | 21:14 |
cdent | true | 21:14 |
cdent | mriedem: I suppose some kind of grenade hoopa joop has to happen soon now there is placement stable/stein? | 21:23 |
mriedem | maybe, https://github.com/openstack-dev/grenade/blob/master/projects/60_nova/from-rocky/upgrade-nova will only run when upgrading nova from rocky to stein, | 21:24 |
mriedem | and in stein devstack we have placement by default | 21:25 |
mriedem | so grenade upgrading from stein to train might just be fine | 21:25 |
cdent | well that will be nice, if so | 21:25 |
mriedem | the stuff that might need to move at some point is https://github.com/openstack-dev/grenade/blob/master/projects/60_nova/resources.sh#L181 | 21:25 |
mriedem | but maybe not | 21:26 |
mriedem | oh now i remember https://review.openstack.org/#/c/631614/ was something that would need to change in grenade with extracted placement | 21:26 |
mriedem | but since it never landed... | 21:26 |
* cdent shrugs | 21:27 | |
cdent | presumably the nova resource checks are something nova continues to care about ("is my inventory still right") | 21:27 |
mriedem | yeah | 21:28 |
openstackgerrit | Chris Dent proposed openstack/placement master: Revert "Centralize registration of logging options" https://review.openstack.org/645014 | 21:32 |
*** e0ne has quit IRC | 22:00 | |
openstackgerrit | Matt Riedemann proposed openstack/placement master: Revert "Centralize registration of logging options" https://review.openstack.org/645014 | 22:10 |
*** mriedem is now known as mriedem_away | 22:25 | |
openstackgerrit | Chris Dent proposed openstack/placement master: Revert "Centralize registration of logging options" https://review.openstack.org/645014 | 22:36 |
openstackgerrit | Chris Dent proposed openstack/placement master: Add register_opts param to PlacementFixture https://review.openstack.org/645031 | 22:36 |
cdent | dammit | 22:36 |
cdent | I -R'd! | 22:36 |
cdent | efried: it's likely I'm gonna fall off the map before that ^ stuff (and the nova side) has fully percolated. Feel free to tweak it and backport to stable/stein (which now exists) as may be required in my absence | 22:41 |
*** mriedem_away has quit IRC | 23:28 | |
*** mriedem has joined #openstack-placement | 23:29 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!