06:02:03 #startmeeting nova api 06:02:04 Meeting started Wed Oct 10 06:02:03 2018 UTC and is due to finish in 60 minutes. The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 06:02:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 06:02:07 The meeting name has been set to 'nova_api' 06:02:38 #link https://wiki.openstack.org/wiki/Meetings/NovaAPI#Agenda_for_next_Office_hours 06:02:41 agenda ^^ 06:03:11 i have not got the much time to review for couple of weeks. 06:03:27 #topic API Subteam Tracking 06:03:34 #linkhttps://etherpad.openstack.org/p/stein-nova-subteam-tracking 06:03:51 i added the approved/under review BP related to API on this extherpad 06:04:01 l63 06:04:36 currently we have an api bp in the runway 06:04:53 yeah 06:05:35 #link https://etherpad.openstack.org/p/nova-runways-stein 06:06:30 yea, the volume type in boot 06:07:09 how much we left for extension merge? 06:07:29 alex_xu: link https://review.openstack.org/#/q/project:openstack/nova+branch:master+topic:bp/api-extensions-merge-stein+status:open 06:07:34 3-4 patch i will say. 06:07:40 i need to rebase them though 06:07:43 cool 06:08:09 this is last patch where i am removing the extensions register logic from wsgi.py - https://review.openstack.org/#/c/607092/ 06:08:26 this is WIP, still mode code to remove 06:09:31 cool, simpler code 06:10:08 yeah, i will work on those and finish soon 06:10:19 thanks 06:12:06 alex_xu: i think we can remove this too - wsgi_action_extensions 06:12:25 do you remember any of action extensions ? 06:13:00 this one - https://github.com/openstack/nova/blob/master/nova/api/openstack/wsgi.py#L468 06:13:04 https://review.openstack.org/#/c/607092/1/nova/api/openstack/compute/routes.py@252 06:13:26 what benefit we remove it? 06:14:56 humm, i was thinking for same resource side but these are all separate entry point for resource action 06:15:25 agree to keep them separate. 06:16:13 still feel a lot of complex thing in https://review.openstack.org/#/c/607092/1/nova/api/openstack/wsgi.py 06:16:50 true, it could be more simpler 06:17:24 without extension, does this still useful https://review.openstack.org/#/c/607092/1/nova/api/openstack/wsgi.py@88 06:17:36 gmann: ^ maybe worth to check this in your last patch 06:17:57 I remember that is used for cache the response obj for the extension, not sure whether it is used by other place 06:18:36 gmann: sorry, I mean those cache interface https://review.openstack.org/#/c/607092/1/nova/api/openstack/wsgi.py@97 06:19:17 yeah those were mainly used in extensions code but i can check if anywhere we use them 06:19:34 but Request object we need 06:20:07 yes, we need Req obj 06:21:46 alex_xu: quickly grep them and it is only extensions code. i will remove them thanks. 06:22:08 cool \o/ 06:22:32 alex_xu: i will make this patch up by tonight and make it ready for you by tomorrow so that u can check if any more bits we can remove 06:22:45 gmann: yea, will do 06:23:39 thanks 06:23:46 moving next 06:24:13 api cleanup things 06:24:15 #link https://etherpad.openstack.org/p/nova-api-cleanup 06:24:47 and spec which need to include the hypervisor API cleanup - https://review.openstack.org/#/c/603969/ 06:25:24 we need to weight on those with what worth to do and what not 06:25:58 and next question on this is - should we do this in stein ? or wait for T to collect more worthy cleanup ? 06:27:02 i can keep it updated for stein and wait till T if any more related API cleanup. And then we do those in single version bump based on agreement 06:27:07 alex_xu: what u say ? 06:27:14 pretty sure we won't get rid of all the cleanup in one microversion 06:27:30 see the past, how much microversion we spend on deprecate proxy API :) 06:27:50 :) yeah 06:28:08 I will review the spec 06:28:13 all you mean "all listed in etherpad" or all in API (because we do not know them now) 06:28:33 all in API 06:28:59 true, so that's is reason i want to wait till T so that we can cover max. 06:29:39 maybe based on the requirement, if there is nothing urgent to change, then we needn't do it very soon 06:29:45 and conclude that "this is we are not going to do" "this is ok to do". so that these does not comes again in future 06:29:52 yeah 06:30:50 i do not think anything urgent on those but i will also get opinion from mriedem and melwitt 06:31:01 also 06:31:20 I'm afriad another option is microversion shouldn't be huge 06:31:35 humm 06:32:18 i feel it should not be huge in term of complexity but should be ok in term of numbers. 06:33:54 anyways that we can discuss on spec with exact list of changes 06:34:20 yea 06:34:38 anything else on this or we move next ? 06:35:19 no more from me 06:35:31 ok 06:35:56 i do not have anything else to discuss form API subteam tracking section. 06:36:02 alex_xu: do you have ? 06:36:13 no 06:36:18 ok 06:36:20 #topic Bug Triage/Discussion 06:36:32 #link https://etherpad.openstack.org/p/nova-api-weekly-bug-report 06:37:16 Tetsuro Nakamura proposed openstack/nova-specs master: Spec: Support filtering by forbidden aggregate https://review.openstack.org/603352 06:38:12 there are 4 new bugs 06:38:57 #link https://bugs.launchpad.net/nova/+bug/1789382 06:38:57 Launchpad bug 1789382 in OpenStack Compute (nova) "openstack server list error" [Undecided,New] 06:39:03 seems timeout from cell ? 06:40:54 also see 'TypeError: 'object' object is not iterable' after cell timeout 06:41:18 cell timeout is just a warning 06:42:32 sounds like we handle something wrong 06:43:08 humm object is none or something may be 06:44:31 let me ask more log of api and compute 06:45:00 i did not find any other pointer to debug without those 06:45:37 yes 06:49:18 done 06:49:28 it seems hitting here - https://github.com/openstack/nova/blob/c6218428e9b29a2c52808ec7d27b4b21aadc0299/nova/compute/multi_cell_list.py#L265 06:50:21 #link https://bugs.launchpad.net/nova/+bug/1796132 06:50:21 Launchpad bug 1796132 in OpenStack Compute (nova) "SSL Verification Error on Launch Instance (queens)" [Undecided,New] 06:50:32 "Seems forcing everythign to https isn't the wisest choice and requires more nova specific ssl config, I'll be going back to using http, so this bug may no longer be valid to work on." 06:50:44 comments from author ^^ 06:52:06 should we mark it invalid then ? 06:52:51 seems like glance was not on https 06:53:06 SSLError: SSL exception connecting to https://10.5.25.19:9292/v2/images/fd9637fb-57e8-4c43-9 06:53:53 I don't know, is it configure mistake? 06:54:11 seems so 06:54:31 i don't familar this part 06:54:40 i will ask on bug is that is solved with all having on https or not 06:56:45 done 06:57:10 it is almost time and other bugs we can do next week or in between of next office hour 06:57:22 yes 06:57:43 #topic Open Discussion 06:57:59 anything else alex_xu you want to discuss otherwise we close for today ? 06:58:08 nothing from me 06:58:24 alex_xu: thanks for joining. this really help. 06:58:32 np 06:58:37 #endmeeting