| *** e0ne has joined #openstack-placement | 05:20 | |
| *** artom has joined #openstack-placement | 05:41 | |
| *** artom is now known as artom|gmtplus3 | 05:41 | |
| *** e0ne has quit IRC | 06:12 | |
| *** belmoreira has joined #openstack-placement | 06:39 | |
| *** ttsiouts has joined #openstack-placement | 06:58 | |
| *** cdent has joined #openstack-placement | 07:04 | |
| *** tssurya has joined #openstack-placement | 07:10 | |
| *** helenafm has joined #openstack-placement | 07:21 | |
| *** ttsiouts has quit IRC | 07:26 | |
| *** ttsiouts has joined #openstack-placement | 07:27 | |
| *** ttsiouts has quit IRC | 07:31 | |
| *** ttsiouts has joined #openstack-placement | 07:51 | |
| *** zzzeek has quit IRC | 08:48 | |
| *** zzzeek has joined #openstack-placement | 08:49 | |
| *** helenafm has quit IRC | 09:00 | |
| *** helenafm has joined #openstack-placement | 09:03 | |
| *** helenafm has quit IRC | 09:15 | |
| *** helenafm has joined #openstack-placement | 09:20 | |
| *** helenafm has quit IRC | 09:34 | |
| *** e0ne has joined #openstack-placement | 09:37 | |
| *** helenafm has joined #openstack-placement | 09:41 | |
| *** lei-zh has joined #openstack-placement | 09:59 | |
| *** lei-zh has quit IRC | 10:01 | |
| *** lei-zh has joined #openstack-placement | 10:02 | |
| *** ttsiouts has quit IRC | 10:16 | |
| *** ttsiouts has joined #openstack-placement | 10:16 | |
| *** ttsiouts has quit IRC | 10:21 | |
| *** lei-zh has quit IRC | 10:36 | |
| *** cdent has quit IRC | 10:58 | |
| *** ttsiouts has joined #openstack-placement | 11:01 | |
| *** tetsuro has joined #openstack-placement | 11:03 | |
| *** cdent has joined #openstack-placement | 11:28 | |
| *** tssurya has quit IRC | 11:34 | |
| *** tssurya has joined #openstack-placement | 11:37 | |
| *** tetsuro has quit IRC | 11:38 | |
| *** tetsuro has joined #openstack-placement | 11:52 | |
| cdent | gibi: if you're around I could use your help in a review of https://review.opendev.org/#/c/666542/ . The way "in" is being used there seems wrong and I'm not communicating the issue well | 12:01 |
|---|---|---|
| *** cdent has quit IRC | 12:08 | |
| *** cdent has joined #openstack-placement | 12:29 | |
| cdent | melwitt: are you gonna be able to come back to https://review.opendev.org/#/c/640898/ eventually ? | 12:39 |
| gibi | cdent: ack. I will try to reach that review today | 12:45 |
| cdent | thanks gibi | 12:45 |
| tssurya | cdent: since SQLite doesn't support addition of constraints to existing table, I am a bit confused on what to do regarding the foreign key I have to add to consumers that will reference the consumer_types_id | 12:56 |
| tssurya | any suggestions ? | 12:56 |
| * cdent thinks | 12:57 | |
| cdent | tssurya: maybe some conditionals like this one: https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/migrate_repo/versions/278_remove_service_fk_in_compute_nodes.py#L27 | 13:01 |
| cdent | (i tried to link to opendev but it was too slow :( ) | 13:02 |
| tssurya | cdent: no problem yea looking into it | 13:02 |
| tssurya | thanks I'll see if I can do something like this | 13:03 |
| *** tetsuro has quit IRC | 13:23 | |
| *** tetsuro has joined #openstack-placement | 13:24 | |
| *** tetsuro has quit IRC | 13:24 | |
| gibi | cdent: I've added my view to https://review.opendev.org/#/c/666542/ . I think there is no solution without some level of end user pain | 13:27 |
| *** ttsiouts has quit IRC | 13:27 | |
| cdent | thanks for that gibi, that makes it more clear | 13:27 |
| cdent | I think your second suggestion is better (and will say so on the review) | 13:28 |
| * cdent doesn't think much in terms of osc-placement... | 13:33 | |
| cdent | I almost always go to curl | 13:33 |
| cdent | but then again, I almost never work with real deployments... | 13:34 |
| *** ttsiouts has joined #openstack-placement | 13:44 | |
| gibi | I use to work with the openstack CLI so it helps me if I don't have to switch to curl for placementy things | 13:47 |
| *** zzzeek has quit IRC | 13:49 | |
| *** zzzeek has joined #openstack-placement | 13:51 | |
| *** zzzeek has quit IRC | 13:53 | |
| *** zzzeek has joined #openstack-placement | 13:54 | |
| *** zzzeek has quit IRC | 13:55 | |
| *** zzzeek has joined #openstack-placement | 13:56 | |
| *** mriedem has joined #openstack-placement | 13:59 | |
| efried | cdent: question | 14:45 |
| cdent | hello | 14:45 |
| efried | can I effectively exceed max_unit by using multiple request groups and group_policy=none? | 14:46 |
| efried | easy enough to test out, but thought you might know off the top | 14:46 |
| efried | more to the point, what would we *want* the behavior to be? | 14:46 |
| efried | If I'm using separate request groups, seems like I want two different "chunks", and having each one (but not the sum) observe max_unit would make sense. | 14:47 |
| cdent | I'd need to look at the code to confirm/deny when the capacity check was happening, but I'd hope/assume it was early, and that yes, you could clobber | 14:47 |
| cdent | and yes | 14:47 |
| efried | I know for sure we do an extra *capacity* check after we consolidate ARRs. | 14:48 |
| efried | but I can't remember if that's for usage only. | 14:48 |
| * efried looks | 14:48 | |
| cdent | yeah, that one check max_unit | 14:49 |
| cdent | line 703 is a_c | 14:49 |
| cdent | s/is/in/ | 14:49 |
| efried | yup | 14:49 |
| cdent | despite what you've said before, that's probably more aligned with the spirit of a allocation | 14:49 |
| cdent | (spirit of which was developed before request groups were ever a thing, so there might be room for wiggle) | 14:50 |
| cdent | but at this point I'd think: let's let it ride unless we need to do something? | 14:50 |
| efried | i.e. "max_unit to represents the most that any one consumer can get out of this resource provider in one allocation" | 14:50 |
| efried | yeah, I can buy that | 14:51 |
| efried | well, this didn't come up out of the blue | 14:51 |
| efried | I'm noodling a situation where it becomes relevant. | 14:51 |
| cdent | " A maximum amount any single allocation against an inventory can have." | 14:51 |
| cdent | (is what the api-ref says) | 14:51 |
| cdent | and begs the question of "what is a single allocation" | 14:51 |
| efried | "payload of a PUT /allocations" | 14:52 |
| efried | If it's that clearly defined in actual documentation, then yeah, let's totally not f with it. | 14:52 |
| cdent | aye | 14:53 |
| melwitt | cdent: yes, been trying to get back to it | 15:01 |
| cdent | melwitt: cool, thank | 15:02 |
| cdent | s | 15:02 |
| cdent | melwitt: there's a similar "how do we name things and make good args" discussion happening on https://review.opendev.org/666542 | 15:02 |
| melwitt | cool, will take a look | 15:03 |
| * cdent goes home | 15:11 | |
| *** tssurya has quit IRC | 15:11 | |
| *** cdent has quit IRC | 15:11 | |
| *** ttsiouts has quit IRC | 15:30 | |
| *** ttsiouts has joined #openstack-placement | 15:31 | |
| *** ttsiouts has quit IRC | 15:36 | |
| *** helenafm has quit IRC | 15:56 | |
| *** e0ne has quit IRC | 16:39 | |
| *** openstackgerrit has quit IRC | 17:04 | |
| *** e0ne has joined #openstack-placement | 17:44 | |
| *** e0ne has quit IRC | 18:06 | |
| *** e0ne has joined #openstack-placement | 18:09 | |
| *** e0ne has quit IRC | 18:22 | |
| *** amodi has quit IRC | 18:25 | |
| *** artom|gmtplus3 has quit IRC | 18:35 | |
| *** e0ne has joined #openstack-placement | 19:39 | |
| *** e0ne has quit IRC | 19:55 | |
| *** e0ne has joined #openstack-placement | 20:20 | |
| *** jaypipes has quit IRC | 20:34 | |
| *** e0ne has quit IRC | 21:28 | |
| *** e0ne has joined #openstack-placement | 21:29 | |
| *** e0ne has quit IRC | 21:46 | |
| *** takashin has joined #openstack-placement | 21:51 | |
| *** efried is now known as efried_pto | 22:45 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!