Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-04-06
15:22:12 bauzas permissive as I can understand crom 'I don't think so' from johnthetubaguy ?
15:22:17 bauzas from*
15:22:32 stephenfin strict for all known namespaces (hw:, os:, vmware:, ...)
15:22:39 stephenfin don't care for anything outside that set
15:24:09 bauzas I'm cool with this
15:24:15 bauzas stephenfin: ^
15:24:30 bauzas no upgrade impact, pretty clear
15:24:47 bauzas stull leaves nothing unchanged except for already-defined namespaces
15:24:50 bauzas that works for me
15:59:35 gibi stephenfin: and if you have out of tree hw:xxx key then add a custom validator?
15:59:46 gibi or use the old microversion
15:59:59 stephenfin yes
16:00:16 gibi so custom validators can add to in-tree name spaces?
16:00:55 stephenfin they shouldn't but they could, yes
16:00:58 gibi OK
16:01:15 stephenfin we just prevent overriding specific extra specs e.g. 'hw:cpu_policy'
16:01:33 gibi OK, cannot override in-tree defined validators
16:01:51 melwitt gmann: we have a minor emergency regarding the policy warning logging (see the -infra channel). it's melting the logstrash indexing
16:02:05 gibi what if I have namespaced out of tree keys today? like ericsson:awesome_feature_flag ?
16:02:28 stephenfin nothing happens until you add a validator
16:02:35 stephenfin we ignore all unknown namespaces
16:02:40 gibi OK, make sense
16:02:50 openstackgerrit John Garbutt proposed openstack/nova master: Update quota_class APIs for db and api limits https://review.opendev.org/712143
16:02:51 stephenfin (and those without namespaces, fwiw)
16:03:12 openstackgerrit John Garbutt proposed openstack/nova master: Add stub unified limits driver https://review.opendev.org/712137
16:03:23 openstackgerrit John Garbutt proposed openstack/nova master: Assert quota related API behavior when noop https://review.opendev.org/712140
16:03:30 gibi stephenfin, johnthetubaguy: seems like a good idea.
16:03:35 openstackgerrit John Garbutt proposed openstack/nova master: Make unified limits APIs return reserved of 0 https://review.opendev.org/712141
16:03:47 openstackgerrit John Garbutt proposed openstack/nova master: Add logic to enforce local api and db limits https://review.opendev.org/712139
16:03:59 openstackgerrit John Garbutt proposed openstack/nova master: Enforce api and db limits https://review.opendev.org/712142
16:04:16 openstackgerrit John Garbutt proposed openstack/nova master: Update quota_class APIs for db and api limits https://review.opendev.org/712143
16:05:31 johnthetubaguy melwitt: oops :/
16:05:43 johnthetubaguy probably means the same for our users
16:05:49 melwitt yeah :( we were just talking about this the other day but I didn't realize how intense it is
16:06:03 johnthetubaguy does it mean almost every API call?
16:06:14 johnthetubaguy or just a massive storm on restart?
16:06:23 melwitt we need to get it removed asap. we were already talking to gmann about it and he was cool with removing it and doing a reno-only announcement re: policy changes
16:06:41 johnthetubaguy melwitt: is there a patch up?
16:06:45 melwitt looks like every API call
16:07:05 johnthetubaguy yeah, that isn't right :/
16:07:46 melwitt no not that I know of, no patch yet
16:14:45 openstackgerrit Alexandre arents proposed openstack/nova master: Calculate disk_over_committed for raw instances https://review.opendev.org/717037
16:14:52 johnthetubaguy its probably all system only API calls... as our tests probably don't have a system admin user yet
16:16:10 openstackgerrit Stephen Finucane proposed openstack/nova master: api: Add microversion for extra spec validation https://review.opendev.org/708436
16:16:11 openstackgerrit Stephen Finucane proposed openstack/nova master: docs: Add documentation for flavor extra specs https://review.opendev.org/710037
16:16:11 openstackgerrit Stephen Finucane proposed openstack/nova master: Drop concept of '?validation' parameter https://review.opendev.org/717789
16:16:12 openstackgerrit Stephen Finucane proposed openstack/nova master: docs: Move description of groups to document itself https://review.opendev.org/717790
16:16:12 openstackgerrit Stephen Finucane proposed openstack/nova master: docs: Add 'nova' domain and include extra specs in it https://review.opendev.org/717791
16:16:45 stephenfin johnthetubaguy: I think that matches up with what you were expecting ^
16:17:06 stephenfin gibi: Also, I retooled the docs. It's _much_ nicer now (we can cross-reference!)
16:17:31 gibi stephenfin: will look but sounds awesome!
16:19:47 openstackgerrit John Garbutt proposed openstack/nova master: Update limit APIs https://review.opendev.org/712707
16:23:56 gibi bauzas: o/ sorry I could not get to the vgpu series of yours
16:24:01 gibi maybe tomorrow
16:24:08 bauzas gibi: thanks and no worries
16:24:33 bauzas gibi: in the meantime, I'll write to provide a func change for seeing the behaviour
16:24:42 openstackgerrit John Garbutt proposed openstack/nova master: Update quota sets APIs https://review.opendev.org/712749
16:24:45 gibi bauzas: thanks
16:29:55 openstackgerrit John Garbutt proposed openstack/nova master: Tell oslo.limit how to count nova resources https://review.opendev.org/713301
16:33:32 openstackgerrit John Garbutt proposed openstack/nova master: Enforce resource limits using oslo.limit https://review.opendev.org/615180
16:40:52 melwitt johnthetubaguy: I wonder if we should revert https://review.opendev.org/701624 for now? or is there some other way to approach this cc gmann
16:41:00 gibi do somebody know the irc nick of the owner of https://review.opendev.org/#/c/713089/ ? we would need to release the novaclient this week and patches are blocked on this
16:43:01 melwitt gibi: looks like it might be alisterle https://launchpad.net/~alistarle
16:43:41 gibi melwitt: thenk I will try to ping him when he is up
16:43:56 johnthetubaguy melwitt: are we sure we know what is causing the logs, I think it might be the scope chagnes
16:44:14 openstackgerrit John Garbutt proposed openstack/nova master: Add legacy limits and usage to unified limits https://review.opendev.org/713498
16:44:14 openstackgerrit John Garbutt proposed openstack/nova master: Update quota apis with keystone limits and usage https://review.opendev.org/713499
16:44:15 openstackgerrit John Garbutt proposed openstack/nova master: Add reno for unified limits https://review.opendev.org/715271
16:45:05 melwitt johnthetubaguy: I don't have deep enough knowledge to know that but I was thinking if we unmarked the things as deprecated it would stop the deprecated messages?
16:45:22 melwitt I guess I would propose the revert as -W to see for sure
16:45:27 johnthetubaguy we have about 20 or more patches like that though
16:45:39 johnthetubaguy do you have the log message causing the issue to hand?
16:45:53 melwitt oh really? it's not just the one that marks deprecated
16:45:57 johnthetubaguy logs I am looking at at failing to load now... which could be related
16:46:04 johnthetubaguy we have deprecated loads though
16:46:10 melwitt ok I see
16:46:34 johnthetubaguy we might have to change the logging config for oslo.policy
16:46:50 johnthetubaguy assuming that is possible somehow
16:47:37 gmann johnthetubaguy: melwitt let me propose it to disable it temp and then find a good way
16:47:51 johnthetubaguy gmann: thanks, sounds like a plan
16:47:59 johnthetubaguy gmann: which log message is it?
16:48:21 melwitt this is an example, https://6d82362f2cdc504b27f1-9f757b11a1d2b00e739d31e1ecad199a.ssl.cf5.rackcdn.com/717662/1/check/tempest-integrated-compute/b3260ce/controller/logs/screen-n-api.txt
16:48:25 gmann johnthetubaguy: it is from base rule
16:49:17 melwitt gmann: yay you're here, thank you. once we disable it I need to tell clarkb so he can dump the current logs from indexing. he said it will never catch up and we need to start fresh
16:55:58 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Disable the policy warning temporary https://review.opendev.org/717802
16:56:18 gmann melwitt: johnthetubaguy ^^
16:58:00 johnthetubaguy gmann: assuming that works, lets do it.
16:59:59 gmann johnthetubaguy: yeah, let's wait for gate result also
17:01:47 melwitt gibi: fyi ^ patch to temporarily disable policy deprecation warnings currently making nova-api logs huge with log spam (example is linked few messages back in backscroll)
17:04:52 johnthetubaguy oh my, its very chatty :/
17:05:17 melwitt yeah it's ... a lot more logging than I had realized
17:09:31 dansmith only 15k in that one log
17:09:56 melwitt "only"
17:09:57 johnthetubaguy ah, right, that isn't so bad then (ducks)
17:10:28 dansmith melwitt: yes, extreme sarcasm around the only
17:10:53 melwitt I know, I thought it was funny
17:11:39 gmann johnthetubaguy: melwitt dansmith : to sync up on some approach on policy warning and adopting new behaviour ( lbragstad and I discussed last week i think)
17:12:22 lbragstad o/
17:12:47 gmann Warning: 1. do not log warning for policy changing defaults (but still not enabled as we support old defaults also) 2. do log warning where policy names are changed (granular cases in our case)
17:13:51 gmann New behaviour: 1. one way was to make enforce_policy as all-new-together a flag to disable the deprecation rules also. means if this flag is true then support only new defaults_scope

Earlier   Later