Earlier  
Posted Nick Remark
#openstack-nova - 2017-09-28
23:02:36 mikal As a man who recently learned those acronymns, isn't that just a session on how to use heat to start instances?
23:05:01 mriedem ha
23:05:04 mriedem maybe?
23:05:10 mriedem but in the etsi way
23:06:37 cdent MANO your VIM with TOSCA
23:07:15 jaypipes dansmith: so after 4 hours fucking around with this, I remember why I made it parent_provider_uuid and root_provider_uuid instead of using internal integer fields.
23:07:28 dansmith jaypipes: lay it on e
23:07:49 mikal cdent: OMG, that sentence even parsed
23:07:57 mikal Why do these people love stupid acronyms so much?
23:08:27 jaypipes dansmith: having to deal with all the friggin relationship stuff in the ORM results in a boatload of code and errors everywhere about lazy-loading and Sessions being inactive..
23:08:27 cdent Don’t acronyms mean importance? I’m sure I read a white paper on that.
23:09:23 dansmith jaypipes: that makes no sense to me
23:09:46 dansmith um, so do it?
23:10:12 jaypipes dansmith: the resource_provider.py module is a mess of mixed up ORM and non-ORM code. drives me nuts.
23:11:52 dansmith mriedem: smackin' the patch now
23:23:43 dansmith mriedem: I got bad news about your sysmeta patch
23:27:53 cdent Goodnight you princes of Maine, you kings of New England.
23:28:08 dansmith mriedem: it _only_ makes our API about 16% faster on average: https://imgur.com/a/ovFq2
23:28:21 mriedem is that bad news?
23:28:36 dansmith terrible
23:28:42 mriedem do i,
23:28:45 mriedem or do i not,
23:28:48 mriedem need a special comb
23:30:32 mriedem crabs
23:30:34 mriedem it's a crabs joke
23:30:40 mriedem you know, bad news
23:30:42 mriedem you've got crabs
23:30:44 dansmith hah, okay
23:32:25 gmann mriedem: yea, tags(all type of query like tag_any, not_tag etc) are in query itself
23:32:51 openstackgerrit Matt Riedemann proposed openstack/nova master: Remove system_metadata loading in Instance._load_flavor https://review.openstack.org/508357
23:33:04 gmann mriedem: api extension policy removal spec is ready - https://review.openstack.org/#/c/508101/2
23:33:32 gmann i listed 11 policies to remove
23:35:50 mriedem gmann: ok, i'm not going to be able to get to that tonight
23:36:01 mriedem we're going to do a spec review sprint early next week
23:36:17 gmann mriedem: sure, btw when is spec freeze date ?
23:36:38 mriedem oct 19
23:37:57 gmann ok. thanks.
23:38:57 dansmith your patch will conflict with my cutover patch, btw
23:40:23 mriedem dansmith: i know, i'll wip
23:40:45 dansmith mriedem: don't have to, just pointing it out
23:41:19 mriedem was just about to go through 505418 and underneath it
23:41:34 mriedem laura is going to have my hide if she comes home at 7 and i don't have dinner started
23:42:03 dansmith I didn't change the cutover one, just put a patch underneath to save the tests
23:42:25 dansmith was going to do your refactoring in the top patch but I haven't yet
23:45:40 mriedem yeah noticed
23:45:43 mriedem the comment i mean
23:45:46 openstackgerrit Steve Noyes proposed openstack/nova master: update live migration to use v3 cinder api https://review.openstack.org/463987
23:47:04 dansmith mriedem: so, I was wondering if maybe doing the cells in parallel would hide this benefit some and it does: https://imgur.com/a/YMxPQ
23:47:38 dansmith that doesn't mean the gain isn't there, it means there's two gains and we're bumping up against just cpu performance processing the results from the DB I think
23:47:41 mriedem the sysmeta benefit?
23:47:41 dansmith which is good
23:47:43 dansmith yeah
23:47:48 mriedem ah
23:47:56 dansmith that's the instance_list patch plus the sysmeta one vs. master
23:48:08 mriedem cool
23:49:39 dansmith I wish we could spawn a real thread instead of a greenthread to do the cells scatter
23:50:37 mriedem https://review.openstack.org/#/c/505418/12
23:50:41 dansmith because we're just pegging one core when we're doing all this
23:50:55 mriedem you know who likes to talk threading?
23:50:58 mriedem that works on your team
23:51:14 mriedem me too
23:51:23 dansmith mriedem_dinner: thanks for the +W
23:51:34 dansmith melwitt: you wanna +W the test refactor underneath that?
23:51:54 dansmith melwitt: it's just copying some tests to the cells deprecated morgue-of-tests so we keep running them in that path
23:52:14 melwitt sure, lemme look
23:52:31 dansmith it would make me the happiest boy in happy valley
23:52:41 melwitt lol
23:57:40 melwitt dansmith: done
23:57:51 cfriesen has anyone else run into problems with devstack when changing branches? I tried going from stable/newton to stable/pike and a bunch of requirements are not properly handled, requiring me to do manual pip stuff.
23:58:03 dansmith melwitt: tanks
#openstack-nova - 2017-09-29
00:30:03 sean-k-mooney dansmith: qq im setting up the live migration environmnet for the multi port binding work
00:30:46 sean-k-mooney devstack complete sucessfully but i dont think i registerd the hyperviors correctly in the toplevel api db http://paste.openstack.org/show/622233/
00:31:18 sean-k-mooney do you know where the doc for how to fix that are i think i have to run a nova-manage command but i cant remember
00:32:54 sean-k-mooney i think i need to run nova-manage cell_v2 discover_hosts?
00:33:09 dansmith sean-k-mooney: discover-hosts?
00:33:09 sean-k-mooney nova-manage cell_v2 list_cells
00:33:17 dansmith nova-manage cell_v2 discover_hosts --verbose
00:33:38 sean-k-mooney nova-manage cell_v2 discover_hosts --verbose
00:33:45 sean-k-mooney wronge terminal
00:33:49 dansmith heh
00:33:55 dansmith er, I mean:
00:34:00 dansmith Password:
00:34:17 sean-k-mooney hehe
00:34:29 sean-k-mooney ya that fixed it
00:34:33 dansmith cool
00:34:49 sean-k-mooney am i guess this might be a bug in devstack then.
00:35:05 dansmith sean-k-mooney: it's just because you have multiple nodes
00:35:13 dansmith devstack-gate does the final discover after both are done in that case
00:35:44 sean-k-mooney yes though i would have expected this to be part the n-cpu service install
00:36:04 dansmith well,
00:36:22 dansmith in our multi-node setups, the second node can't see the database on purpose, so it can't do that
00:36:37 dansmith but I guess you could make it conditional in devstack so it was graceful
00:36:42 sean-k-mooney oh ya right so it has to run on the controler
00:36:48 dansmith yeah
00:37:21 sean-k-mooney ok that makes sense well i know for the future havent done multi node with devstack in a while
00:37:35 sean-k-mooney thanks for the help
00:46:09 dansmith np
01:14:27 openstackgerrit Takashi NATSUME proposed openstack/nova master: List/show all server migration types (1/2) https://review.openstack.org/430608
01:18:45 openstackgerrit Naichuan Sun proposed openstack/nova master: xenapi: Support live migration in pooled multi-nodes environment https://review.openstack.org/489451
01:19:35 openstackgerrit Takashi NATSUME proposed openstack/nova master: List/show all server migration types (1/2) https://review.openstack.org/430608

Earlier   Later