| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-03-10 | |||
| 13:57:57 | openstackgerrit | Ildiko Vancsa proposed openstack/nova master: Deprecate the Cinder API v2 support https://review.openstack.org/444253 | |
| 14:15:35 | mriedem | o/ | |
| 14:18:02 | Dinesh_Bhor | mriedem: Hi, if you get time could you please check this https://review.openstack.org/#/c/441750/? | |
| 14:35:14 | mriedem | dansmith: i think we're happy with this now https://review.openstack.org/#/c/437222/ - the notification delete patch | |
| 14:40:58 | ioni | hello guys | |
| 14:41:22 | ioni | sorry for bothering you, i have noticed something and I don't know if is a bug in nova or a bug in documentation | |
| 14:41:30 | ioni | https://docs.openstack.org/ocata/config-reference/compute/config-options.html | |
| 14:41:44 | ioni | is mentioned that notify_on_state_change is in [DEFAULT] | |
| 14:42:11 | ioni | while looking at the logs, i have noticed that notifications.notify_on_state_change = None | |
| 14:42:44 | ioni | the only way to set this up is to create a new [notifications] section, having notify_on_state_change = vm_and_task_state | |
| 14:42:57 | ioni | bug in nova or bug in documentation? | |
| 14:45:27 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/nova-specs master: Enable SR-IOV NIC offload feature discovery https://review.openstack.org/435954 | |
| 14:47:32 | openstackgerrit | Matt Riedemann proposed openstack/nova-specs master: Websockify security proxy framework https://review.openstack.org/443234 | |
| 14:54:59 | figleaf | ioni: that's a doc bug | |
| 14:55:21 | figleaf | are you running ocata? | |
| 14:55:45 | ioni | figleaf, yes | |
| 14:57:04 | figleaf | ioni: https://github.com/openstack/nova/blob/master/nova/conf/notifications.py#L108-L110 | |
| 14:57:39 | figleaf | ioni: that shows the opts defined above it, including notify_state_on_change, as being part of the notifications_group | |
| 14:57:41 | mriedem | ioni: are you sure that it wasn't deprecated in the DEFAULT group in ocata and still pass through to [notifications]? | |
| 14:57:59 | mriedem | we generally deprecate the group for a release when moving them | |
| 14:58:04 | mriedem | so the old config will work | |
| 14:58:11 | mriedem | so it depends on when using it in DEFAULT was deprecated | |
| 14:58:29 | ioni | mriedem, i'm looking at: https://docs.openstack.org/ocata/config-reference/compute/config-options.html | |
| 14:58:38 | mriedem | https://docs.openstack.org/developer/nova/ocata/sample_config.html | |
| 14:58:46 | mriedem | yeah, the docs.o.o might be wrong | |
| 14:58:50 | mriedem | ^ is generated from the nova repo | |
| 14:58:55 | ioni | i don't know what's in the code, i'm just looking at them :D | |
| 14:59:08 | mriedem | Deprecated group/name - [default]/notify_on_state_change | |
| 14:59:37 | ioni | ok | |
| 14:59:42 | mriedem | https://github.com/openstack/nova/blob/stable/ocata/nova/conf/notifications.py#L33 | |
| 14:59:52 | mriedem | so if you have it in [DEFAULT] in ocata it should still work | |
| 14:59:55 | mriedem | you might get a warning | |
| 15:00:01 | ioni | it doesn't work | |
| 15:00:08 | ioni | no warning, it defaults to None | |
| 15:00:08 | mriedem | sfinucan: does deprecated_group='default' work for [DEFAULT]? | |
| 15:00:19 | mriedem | dims_: ^? | |
| 15:00:20 | openstackgerrit | Dan Smith proposed openstack/nova master: Get instance availability_zone without hitting the api db https://review.openstack.org/439754 | |
| 15:00:20 | openstackgerrit | Dan Smith proposed openstack/nova master: Make os-availability-zones know about cells https://review.openstack.org/442163 | |
| 15:00:21 | openstackgerrit | Dan Smith proposed openstack/nova master: Change MQ targeting to honor only what is in the context https://review.openstack.org/438023 | |
| 15:00:22 | openstackgerrit | Dan Smith proposed openstack/nova master: Make scheduler target cells to get compute node instance info https://review.openstack.org/439891 | |
| 15:00:23 | openstackgerrit | Dan Smith proposed openstack/nova master: Add identity helper property to CellMapping https://review.openstack.org/444038 | |
| 15:00:24 | openstackgerrit | Dan Smith proposed openstack/nova master: Add cell field to Destination object https://review.openstack.org/438024 | |
| 15:00:24 | openstackgerrit | Dan Smith proposed openstack/nova master: Make conductor ask scheduler to limit migrates to same cell https://review.openstack.org/438025 | |
| 15:01:09 | mriedem | ioni: let me try something quick in ocata | |
| 15:01:43 | sfinucan | mriedem: I think macsz tried that and it does | |
| 15:02:45 | mriedem | ok i'm changing a unit test in ocata that sets it into the 'notifications' group, putting it back in DEFAULT and seeing if that passes | |
| 15:03:14 | mriedem | oslo_config.cfg.NoSuchOptError: no such option notify_on_state_change in group [DEFAULT] | |
| 15:04:53 | mriedem | changing deprecated_group="DEFAULT" | |
| 15:05:33 | ioni | mriedem, https://paste.xinu.at/VmpI/ | |
| 15:05:58 | ioni | mriedem, so in the first two lines, i had notify_on_state_change in [DEFAULT] | |
| 15:06:08 | ioni | no warning, nothing related to that | |
| 15:06:42 | mriedem | this doesn't work for me either https://gist.github.com/mriedem/d20978819dbbc28c77f89e118fd4f52f | |
| 15:06:52 | mriedem | i get oslo_config.cfg.NoSuchOptError: no such option notify_on_state_change in group [DEFAULT] | |
| 15:06:55 | sfinucan | mriedem: AFAIK the deprecated_group and deprecated_name properties only affects reading from config files. When accessing it programatically (cfg.CONF) you need to use the current group and name | |
| 15:06:57 | mriedem | might be a problem in the unit tests | |
| 15:07:04 | mriedem | sfinucan: ah | |
| 15:07:29 | sfinucan | which, imo, is the correct behavior | |
| 15:07:32 | mriedem | ioni: are you able to change nova/conf/notifications.py and change deprecated_group='DEFAULT', restart and see if that works? | |
| 15:07:48 | mriedem | i don't have a devstack in front of me to recreate | |
| 15:08:06 | ioni | ok, wait | |
| 15:08:21 | dims_ | mriedem : sfinucan : we just merged something that may help in the future - https://review.openstack.org/#/c/357987/ | |
| 15:10:38 | ioni | mriedem, deprecated_group='default' this line? | |
| 15:10:48 | mriedem | ioni: yeah | |
| 15:10:51 | ioni | with capital DEFAULT | |
| 15:10:51 | ioni | right? | |
| 15:10:53 | mriedem | yes | |
| 15:11:37 | ioni | 2017-03-10 17:11:26.849 14528 WARNING oslo_config.cfg [req-13263cac-0dca-42bd-9c0a-d2e3cbab71a6 - - - - -] Option "notify_on_state_change" from group "DEFAULT" is deprecated. Use option "notify_on_state_change" from group "notifications". | |
| 15:11:39 | ioni | now it works | |
| 15:12:15 | mriedem | ioni: ok, crap, sorry - can you report a bug against nova? | |
| 15:12:20 | ioni | should i submit a patch ? | |
| 15:12:24 | mriedem | https://bugs.launchpad.net/nova | |
| 15:12:40 | mriedem | ioni: sure, i see we have the same mistake for some console options | |
| 15:12:54 | mriedem | ioni: report a bug, submit a patch and then we can work on reviewing and backporting to ocata | |
| 15:13:00 | mriedem | i'm going to do a stable/ocata release next week, we can get this fix in there | |
| 15:13:21 | mriedem | dims_: ^ seems like a bug in oslo.config if you can specify a deprecated_group that doesn't exist | |
| 15:13:49 | mriedem | maybe that's a chicken and egg, | |
| 15:13:58 | mriedem | since the deprecated group is no longer registered (except the DEFAULT case) | |
| 15:14:57 | mriedem | johnthetubaguy: were you going to re-propose a spec for pike for this? https://review.openstack.org/#/c/393888/ | |
| 15:15:10 | dims_ | mriedem : a bit distracted at the moment. bug please? | |
| 15:15:35 | mriedem | forthcoming, i'll point it out when it's up | |
| 15:16:30 | sfinucan | mriedem: Yeah, I reviewed that patch and missed that. Sorry :( | |
| 15:16:44 | dims_ | thanks mriedem ! | |
| 15:16:51 | sfinucan | I'll try see if we can fix this in oslo.config, if that's ok with you dims_ | |
| 15:17:27 | openstack | Launchpad bug 1671847 in OpenStack Compute (nova) "Incorect set deprecated flag for notify_on_state_change" [Undecided,New] | |
| 15:17:27 | ioni | mriedem, report here: https://bugs.launchpad.net/nova/+bug/1671847 | |
| 15:17:32 | ioni | mriedem, patch coming up | |
| 15:18:08 | sfinucan | tonyb, lyarwood: Could ye stick this on your review queues? https://review.openstack.org/#/c/444306/ | |
| 15:18:41 | sfinucan | I'm seeing the issue in openstack-manuals for stable/ocata still. Not a huge problem but it's very annoying | |
| 15:18:55 | cdent | leakypipes: is https://review.openstack.org/#/c/442718/ -w cuz the tests aren't done, or are you reconsidering the idea? (I'm gathering things for this week's updated) | |
| 15:19:52 | cdent | counting number of available mempages seems ... odd? | |
| 15:19:55 | mriedem | ioni: thanks | |
| 15:21:01 | mriedem | cdent: also, someone needs to re-propose https://review.openstack.org/#/c/386710/ for pike and update it to mention the discussion at the ptg about not supporting multiple parents, | |
| 15:21:20 | cdent | mriedem: ✔ | |
| 15:21:21 | mriedem | doesn't need great detail, but that seemed like a long whiteboarding session and if that was the outcome we should probably capture in the spec with the reason why | |
| 15:22:07 | lyarwood | sfinucan: ack, looking | |
| 15:23:14 | sfinucan | lyarwood: (y) | |
| 15:25:05 | openstackgerrit | Ionuț Bîru proposed openstack/nova master: Fix notify_on_state_change deprecation https://review.openstack.org/444350 | |
| 15:25:22 | ioni | mriedem, ^ | |
| 15:27:26 | dims_ | ioni : does it work if you just delete the line 33? (https://review.openstack.org/#/c/444350/1/nova/conf/notifications.py) | |
| 15:28:05 | ioni | testing now | |