| *** mriedem has quit IRC | 00:45 | |
| *** tetsuro has joined #openstack-placement | 02:20 | |
| *** dklyle has quit IRC | 04:00 | |
| *** david-lyle has joined #openstack-placement | 04:00 | |
| *** tetsuro has quit IRC | 05:12 | |
| *** tetsuro has joined #openstack-placement | 05:18 | |
| *** e0ne has joined #openstack-placement | 05:19 | |
| *** e0ne has quit IRC | 05:20 | |
| *** e0ne has joined #openstack-placement | 05:51 | |
| *** e0ne has quit IRC | 06:24 | |
| *** e0ne has joined #openstack-placement | 06:39 | |
| openstackgerrit | Tetsuro Nakamura proposed openstack/placement master: Refactor initialization of research context https://review.opendev.org/679385 | 07:32 |
|---|---|---|
| *** cdent has joined #openstack-placement | 07:39 | |
| *** tetsuro has quit IRC | 07:46 | |
| *** cdent has quit IRC | 08:06 | |
| *** e0ne has quit IRC | 08:40 | |
| *** e0ne has joined #openstack-placement | 08:40 | |
| *** cdent has joined #openstack-placement | 08:41 | |
| *** e0ne has quit IRC | 08:46 | |
| *** e0ne has joined #openstack-placement | 09:03 | |
| *** e0ne has quit IRC | 10:01 | |
| *** e0ne has joined #openstack-placement | 10:10 | |
| *** efried has quit IRC | 10:24 | |
| *** irclogbot_1 has quit IRC | 10:24 | |
| *** edleafe has quit IRC | 10:24 | |
| *** evrardjp has quit IRC | 10:24 | |
| *** efried has joined #openstack-placement | 10:27 | |
| *** evrardjp has joined #openstack-placement | 10:27 | |
| *** irclogbot_1 has joined #openstack-placement | 10:28 | |
| *** artom has quit IRC | 10:58 | |
| *** artom has joined #openstack-placement | 10:59 | |
| *** artom has quit IRC | 11:00 | |
| *** artom has joined #openstack-placement | 11:01 | |
| *** artom has quit IRC | 11:01 | |
| *** artom has joined #openstack-placement | 11:01 | |
| *** artom has quit IRC | 11:01 | |
| *** artom has joined #openstack-placement | 11:22 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Microversion 1.37: Adds support for consumer_types https://review.opendev.org/672476 | 11:29 |
| cdent | biab | 11:30 |
| *** cdent has quit IRC | 11:30 | |
| *** cdent has joined #openstack-placement | 11:52 | |
| efried | cdent: You got a minute to help me rebuild a busted placement db? | 12:10 |
| cdent | efried: i can try, what's up? | 12:10 |
| efried | cdent: had a devstack | 12:10 |
| efried | it was working, left it alone a few days, came back (it's possible dustinc did some stuff with it) and journalctl was saying "couldn't find database 'placement'" | 12:11 |
| efried | the service was still alive | 12:11 |
| efried | but the logs end with | 12:11 |
| efried | *** no app loaded. going in full dynamic mode *** | 12:11 |
| efried | and requests 500 | 12:12 |
| efried | So I tried rebuilding manually by stealing the mysql commands from the perfload job | 12:12 |
| efried | but now on startup it does stuff like | 12:12 |
| efried | ProgrammingError: (pymysql.err.ProgrammingError) (1146, u"Table 'placement.traits' doesn't exist") | 12:12 |
| cdent | hmm | 12:12 |
| efried | so I tried setting the sync_on_startup thing | 12:13 |
| efried | and it would complain about tables already existing | 12:13 |
| efried | (don't have that one at my fingertips, can get it for you if you need) | 12:13 |
| cdent | what outcome do you want: a) the data you had before, b) some placement db with lots of stuff in it ? | 12:14 |
| efried | c) a fresh placement db like if I just devstacked, so nova will repopulate it. | 12:14 |
| efried | last resort I can restack. | 12:15 |
| efried | but thought if there was an easier way... | 12:15 |
| efried | ("easier" for you obviously. for me...) | 12:15 |
| efried | cdent: This system is publicly accessible if you want to get on and play | 12:16 |
| cdent | drop the database, create the database, and either let the sync on startup do its thing or do db_sync, restart nova-compute | 12:17 |
| efried | yeah, that's what I've been doing. | 12:18 |
| cdent | the fact that those drop and creates commands are obscure things I can't remember is why I tend to use postgres (dropdb placement; createdb placement) | 12:18 |
| efried | I don't get as far as 'restart nova-compute' because the startup spews those errors. | 12:18 |
| cdent | if those things aren't working, make sure there isn't a mismatch between your database connection in placement.conf and what you're doing command line wise | 12:18 |
| efried | doesn't seem to be, because I would think it wouldn't be getting as far as it is (also I checked visually) | 12:19 |
| cdent | if you want to paste me the access bits I can do a quick look | 12:20 |
| *** mriedem has joined #openstack-placement | 12:51 | |
| *** e0ne has quit IRC | 13:05 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add consumer_types table and consumer_type_id column to consumers https://review.opendev.org/669170 | 13:09 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Consumer_types: the database and object bits https://review.opendev.org/679440 | 13:09 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Microversion 1.37: API support for consumer types https://review.opendev.org/679441 | 13:09 |
| *** edleafe has joined #openstack-placement | 13:11 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Consumer_types: the database and object bits https://review.opendev.org/679440 | 13:11 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Microversion 1.37: API support for consumer types https://review.opendev.org/679441 | 13:11 |
| *** cdent has quit IRC | 13:11 | |
| *** e0ne has joined #openstack-placement | 13:14 | |
| *** edmondsw_ has joined #openstack-placement | 13:31 | |
| *** edmondsw_ has quit IRC | 13:41 | |
| *** cdent has joined #openstack-placement | 13:55 | |
| *** e0ne has quit IRC | 14:35 | |
| *** e0ne has joined #openstack-placement | 14:39 | |
| *** e0ne has quit IRC | 16:04 | |
| openstackgerrit | Merged openstack/osc-placement master: Add --aggregate option to 'resource provider inventory set' https://review.opendev.org/640898 | 16:05 |
| *** e0ne has joined #openstack-placement | 16:09 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Consumer_types: the database and object bits https://review.opendev.org/679440 | 16:14 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Microversion 1.37: API support for consumer types https://review.opendev.org/679441 | 16:14 |
| *** mriedem is now known as mriedem_afk | 16:16 | |
| cdent | efried, mriedem : in the absence of melwitt and tssurya that stuff is close to ready for looksies by other folk to have it ready for nova by U | 16:16 |
| efried | hey cdent, qq on api-ref | 16:18 |
| efried | For a given microversion, we always had the format returned by GET /a_c match up with PUT /allocations, right? | 16:18 |
| efried | And at some early point that was a listy format, right? | 16:18 |
| efried | I was looking yesterday and didn't see where PUT /allocs could take the listy format. | 16:19 |
| efried | GET /a_c was a 1.10 (and possibly 1.11, can't exactly remember) | 16:19 |
| cdent | i need to pull something out of my brains | 16:19 |
| cdent | from tape | 16:20 |
| efried | It's only the anal part of me that cares; nobody should be using those old microversions anyway. | 16:20 |
| efried | But the anal part says "IT'S GOTTA BE RIGHT!" | 16:20 |
| cdent | okay | 16:24 |
| cdent | it's messy | 16:24 |
| cdent | pre 1.12 you could PUT allocations as a list | 16:25 |
| cdent | so in 1.10 - 1.12 of a_c allocations was a list of dictionaries | 16:26 |
| cdent | sorry, 1.10 and 1.11 | 16:26 |
| cdent | at 1.12 we got our minds right and made it "A dictionary of allocations keyed by resource provider uuid." | 16:26 |
| cdent | so PUT /alloc from 1.0 - 1.11 was listy, but GET was dicty | 16:27 |
| cdent | which was weird | 16:27 |
| cdent | and thankfully we fixed it in the dicty direction | 16:28 |
| cdent | because lists are trouble | 16:28 |
| cdent | efried: does your butt feel better now? | 16:28 |
| efried | cdent: that's what I remember of history, but when I looked at the api-ref yesterday it didn't seem to reflect that. | 16:29 |
| efried | let me pull it up again and see if it's changed since I slept... | 16:29 |
| cdent | it does, it's just hard to fine because there's too much data in the a_c description | 16:29 |
| cdent | s/data/examples/\ | 16:29 |
| cdent | you have to scroll and scroll and hunt and search | 16:29 |
| cdent | and then you explode | 16:29 |
| cdent | and someone asks for bitcoins | 16:30 |
| *** e0ne has quit IRC | 16:30 | |
| efried | okay, I see it now, not sure what I was looking at yesterday, maybe POST /allocations, which didn't exist at all prior to 1.13. | 16:31 |
| efried | butt is fine | 16:31 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Fix typo in microversion sequence test https://review.opendev.org/679480 | 16:34 |
| cdent | word | 16:34 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Switch ConsumerType to use an AttributeCache https://review.opendev.org/679486 | 17:13 |
| cdent | there we go, got next week's work done today... | 17:15 |
| cdent | have a good weekend | 17:15 |
| * cdent waves | 17:15 | |
| *** cdent has quit IRC | 17:15 | |
| *** mriedem_afk is now known as mriedem | 19:24 | |
| openstackgerrit | Merged openstack/placement master: Fix typo in microversion sequence test https://review.opendev.org/679480 | 19:43 |
| *** mriedem has quit IRC | 22:09 | |
| *** artom has quit IRC | 22:53 | |
| *** N3l1x has quit IRC | 23:44 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!