| *** mriedem has joined #openstack-placement | 00:17 | |
| *** mriedem has quit IRC | 00:30 | |
| *** sean-k-mooney has quit IRC | 02:28 | |
| *** sean-k-mooney has joined #openstack-placement | 03:00 | |
| *** sean-k-mooney has quit IRC | 03:14 | |
| *** sean-k-mooney has joined #openstack-placement | 03:16 | |
| *** sean-k-mooney has quit IRC | 04:19 | |
| *** sean-k-mooney has joined #openstack-placement | 04:21 | |
| *** bhagyashris has joined #openstack-placement | 06:19 | |
| *** tssurya has joined #openstack-placement | 07:07 | |
| *** ttsiouts has joined #openstack-placement | 07:08 | |
| *** ttsiouts has quit IRC | 07:25 | |
| *** ttsiouts has joined #openstack-placement | 07:26 | |
| *** ttsiouts_ has joined #openstack-placement | 07:28 | |
| *** ttsiouts has quit IRC | 07:30 | |
| *** ttsiouts_ has quit IRC | 07:38 | |
| *** mcgigglier has joined #openstack-placement | 07:38 | |
| *** ttsiouts has joined #openstack-placement | 07:39 | |
| *** ttsiouts has quit IRC | 07:43 | |
| *** ttsiouts has joined #openstack-placement | 08:00 | |
| *** takashin has left #openstack-placement | 08:02 | |
| *** e0ne has joined #openstack-placement | 08:23 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: perfload with written allocations https://review.opendev.org/660754 | 09:17 |
|---|---|---|
| openstackgerrit | Chris Dent proposed openstack/placement master: DNM: See what happens with 10000 resource providers https://review.opendev.org/657423 | 09:18 |
| *** cdent has joined #openstack-placement | 09:20 | |
| gibi | cdent: hi! do you happen to know if standalone placement supports osprofiler? I found some imports in the code but I'm not sure the support is complete | 09:28 |
| cdent | gibi: I don't know what it was every completed, either in nova or out | 09:28 |
| cdent | It's certainly never been tested (that I'm aware of) beyond the original addition of the imports | 09:29 |
| gibi | cdent: OK. I guess I have to try to use it and see if it works | 09:29 |
| gibi | the nova side seems to work pretty well already in Pike | 09:29 |
| cdent | yeah, that sounds like the right plan | 09:29 |
| cdent | nothing has changed without how it is being imported, so if it used to work, it ought to still | 09:30 |
| cdent | osprofiler is one of those things I've never known anything about | 09:30 |
| cdent | gibi: if it is working (or you get it working) we should probably write down somewhere in the docs that it does and explain what it is good for | 09:32 |
| gibi | cdent: sure. I will report back when I manage to handle this newest TODO of mine | 09:33 |
| gibi | I think the sentences in the doc https://docs.openstack.org/osprofiler/latest/ defines at least what osprofiler is good for | 09:34 |
| cdent | Yeah, I just read that. The key seems to be the cross-project interaction | 09:35 |
| gibi | cdent: we was able to trace timing interaction between nova-conductor - nova-scheduler - db | 09:37 |
| cdent | so I guess "cross-service" is more accurate | 09:37 |
| gibi | yes | 09:38 |
| gibi | and I noticed that in this env (which is on Pike) there is no placement trace | 09:38 |
| gibi | hence my original question | 09:38 |
| cdent | it looks like in placement it is turned on in wsgi.py which means that it won't show up in placement functional tests, or in nova's use of the placement fixture | 09:41 |
| cdent | there are also no trace points within placement itself | 09:42 |
| gibi | I think it would be nice to see the db interactions from placement as well | 09:46 |
| cdent | as you well know, they are not easily contained | 09:48 |
| cdent | I guess you could trace from the interface betwen the handlers and the "objects" code | 09:49 |
| *** bhagyashris has quit IRC | 09:55 | |
| cdent | gibi: I've found it more informative to distinguish at the service level and then once you decide "placement is not as fast as I would like for this kind of request" to use https://review.opendev.org/#/c/643269/ to get much more detail info (as its the details that usually count) | 09:56 |
| *** ttsiouts has quit IRC | 10:24 | |
| *** ttsiouts has joined #openstack-placement | 10:25 | |
| *** ttsiouts has quit IRC | 10:29 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Stabilize AllocationRequest hash https://review.opendev.org/663137 | 10:47 |
| cdent | efried: I added a few more tests to ^ to make sure we're all on the same page on what we think the results _should_ be | 10:49 |
| cdent | gibi: you might find that interesting | 10:49 |
| *** ttsiouts has joined #openstack-placement | 10:57 | |
| *** cdent has quit IRC | 11:00 | |
| *** jaypipes has joined #openstack-placement | 11:01 | |
| *** ttsiouts has quit IRC | 11:01 | |
| *** ttsiouts has joined #openstack-placement | 11:01 | |
| *** cdent has joined #openstack-placement | 11:01 | |
| gibi | cdent: thanks for the pointer | 11:35 |
| *** dansmith has quit IRC | 12:29 | |
| *** dansmith has joined #openstack-placement | 12:32 | |
| *** artom has quit IRC | 12:35 | |
| *** ttsiouts has quit IRC | 13:01 | |
| cdent | bbl | 13:06 |
| *** cdent has quit IRC | 13:06 | |
| *** ttsiouts has joined #openstack-placement | 13:08 | |
| *** mriedem has joined #openstack-placement | 13:10 | |
| openstackgerrit | Merged openstack/placement master: Bump os-traits to latest release (0.14.0) https://review.opendev.org/663735 | 13:22 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Stabilize AllocationRequest hash https://review.opendev.org/663137 | 13:30 |
| *** artom has joined #openstack-placement | 13:42 | |
| openstackgerrit | Balazs Gibizer proposed openstack/placement master: Add support for osprofiler in wsgi https://review.opendev.org/663945 | 13:44 |
| openstackgerrit | Balazs Gibizer proposed openstack/placement master: Add support for osprofiler in wsgi https://review.opendev.org/663945 | 14:15 |
| *** cdent has joined #openstack-placement | 14:21 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: perfload with written allocations https://review.opendev.org/660754 | 14:27 |
| efried | cdent: I don't understand what would need to happen at release time in order for things in required-projects to become real releases. Wouldn't we need to have a big and confusing flurry of constraints updates, all at the same time, while we're trying to get other release stuff done, with potential chicken/egg problems? | 14:32 |
| cdent | I don't know. That's why I keep asking. If neither of us know, then that's fine, we can just leave it be and do the small amount of work we've been doing. | 14:33 |
| efried | Maybe we should have this conversation in -infra, or -releases, or something. | 14:33 |
| cdent | I've never been a fan of pinned dependencies | 14:33 |
| efried | because I would like to understand it as well. | 14:33 |
| efried | because ever since we moved to this system, it seems like depends-on has never worked the way I want it to. | 14:33 |
| cdent | switching to independent release didn't impact our depencies on os-traits | 14:34 |
| efried | I mean zuulv3, not independent release model for os-traits/-resource-classes. | 14:34 |
| efried | not sure if it was actually zuulv3 that made things different/weird, but it was roughly around that time | 14:35 |
| cdent | if you want to spin something up in -releases or -infra i'm in both right now, but I'm going to struggle to articulate the points | 14:35 |
| cdent | (and usual "hey, what about email?" blahs apply) | 14:35 |
| efried | For my part, I'm not sure I care enough. | 14:35 |
| efried | Maybe some day it will get painful enough that I'll feel like chasing it down | 14:36 |
| efried | but if that hasn't happened yet (I think I've had four series in the past few weeks that have suffered from this) it probably ain't gonna. | 14:36 |
| cdent | i'm not sure I care enough either, so we can let it ride for a while | 14:38 |
| cdent | it just feels dumb | 14:39 |
| cdent | and at some point it would be nice to make it less dumb | 14:39 |
| gibi | cdent: thanks for the pointer in the osprofiler patch. Exactly the type of help I was needed. | 14:43 |
| cdent | you're welcome | 14:43 |
| cdent | I suspect the difference in moving parts is part of why it was never finished in placement | 14:43 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Stabilize AllocationRequest hash https://review.opendev.org/663137 | 15:00 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add NUMANetworkFixture for gabbits https://review.opendev.org/657463 | 15:00 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Prepare objects for allocation request mappings https://review.opendev.org/662785 | 15:00 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Stabilize AllocationRequest hash https://review.opendev.org/663137 | 15:01 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add NUMANetworkFixture for gabbits https://review.opendev.org/657463 | 15:01 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Prepare objects for allocation request mappings https://review.opendev.org/662785 | 15:02 |
| cdent | bbs | 15:08 |
| *** cdent has quit IRC | 15:08 | |
| *** e0ne has quit IRC | 15:22 | |
| *** mcgigglier has quit IRC | 15:29 | |
| *** ttsiouts has quit IRC | 15:38 | |
| *** ttsiouts has joined #openstack-placement | 15:39 | |
| *** cdent has joined #openstack-placement | 15:39 | |
| *** gibi is now known as gibi_ogg | 15:41 | |
| *** gibi_ogg is now known as gibi_off | 15:41 | |
| *** ttsiouts has quit IRC | 15:43 | |
| *** stephenfin is now known as finucannot | 16:05 | |
| *** mriedem has left #openstack-placement | 16:28 | |
| *** mriedem has joined #openstack-placement | 16:29 | |
| *** Sundar has joined #openstack-placement | 16:30 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add NUMANetworkFixture for gabbits https://review.opendev.org/657463 | 16:39 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Prepare objects for allocation request mappings https://review.opendev.org/662785 | 16:39 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Implement allocation candidate mappings https://review.opendev.org/662245 | 16:39 |
| cdent | efried: I fixed a buglet in the NUMANetworkFixture ^ and got those next two working properly with the new hashing and sorting. The latter still needs allocation schema changes. | 16:43 |
| cdent | and that's me done | 16:43 |
| cdent | have a good weekend everyone | 16:43 |
| *** cdent has quit IRC | 16:43 | |
| *** efried is now known as fried_rolls | 16:51 | |
| *** gibi_off has quit IRC | 16:52 | |
| *** tssurya has quit IRC | 17:36 | |
| *** Sundar has quit IRC | 17:44 | |
| *** finucannot is now known as stephenfin | 18:02 | |
| *** e0ne has joined #openstack-placement | 18:59 | |
| *** Sundar has joined #openstack-placement | 19:03 | |
| *** fried_rolls is now known as efried | 19:15 | |
| *** e0ne has quit IRC | 19:24 | |
| *** e0ne has joined #openstack-placement | 19:36 | |
| *** Sundar has quit IRC | 19:44 | |
| *** mriedem has quit IRC | 20:08 | |
| *** mriedem has joined #openstack-placement | 20:11 | |
| *** Sundar has joined #openstack-placement | 20:31 | |
| *** e0ne has quit IRC | 20:37 | |
| *** artom has quit IRC | 20:39 | |
| *** cfriesen has quit IRC | 20:54 | |
| *** mriedem has quit IRC | 21:06 | |
| *** e0ne has joined #openstack-placement | 21:31 | |
| *** e0ne has quit IRC | 21:40 | |
| *** Sundar has quit IRC | 22:16 | |
| *** artom has joined #openstack-placement | 23:05 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!