| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-01-27 | |||
| 01:02:36 | mriedem | for https://launchpad.net/bugs/1658571 we said we'd fix it after FF as a bug fix, but that was before i was reminded that we only have 1 week until rc1 | |
| 01:02:36 | openstack | Launchpad bug 1658571 in OpenStack Compute (nova) "Microversion 2.37 break 2.32 usage" [High,In progress] - Assigned to Artom Lifshitz (notartom) | |
| 01:03:01 | gmann | mriedem: +1 | |
| 01:03:10 | alex_xu | mriedem: 1 week from now? | |
| 01:03:15 | mriedem | alex_xu: yeah | |
| 01:03:25 | alex_xu | ok, got it | |
| 01:03:29 | mriedem | so, that one might end up waiting until pike | |
| 01:03:34 | mriedem | since it was already broken in newton | |
| 01:03:40 | mriedem | we might just have to document it for ocata | |
| 01:03:56 | alex_xu | ok | |
| 01:04:22 | mriedem | i've tagged the bug with ocata-rc-potential though so we don't forget about it | |
| 01:04:30 | alex_xu | cool | |
| 01:04:36 | gmann | it looks pretty close if we can get in Ocata only but yea let's not hurry on that may be | |
| 01:04:48 | alex_xu | gmann: +1 | |
| 01:05:02 | gmann | alex_xu: but for old deployer timestamp will not be changed right? | |
| 01:05:25 | gmann | alex_xu: where we can see those warning so that we can verify the same | |
| 01:05:59 | alex_xu | gmann: emm...they should get a warning after upgrade, due to 'saved_file_rules' is empty | |
| 01:07:00 | gmann | alex_xu: ah yea, thats is nice point | |
| 01:16:32 | artom | mriedem, is one week really too little? | |
| 01:16:51 | artom | I've addressed gmann's and alex_xu's feedback :) | |
| 01:17:14 | gmann | artom: +1, ll try to check today. m sure that is very close | |
| 01:17:52 | artom | The patch itself is fairly straightforward, if discussion is needed I think it's more about the strategy | |
| 01:18:04 | mriedem | artom: w/o looking at the change i can't really say | |
| 01:18:12 | mriedem | artom: we should have a spec for the thing too since it's a new microversion right? | |
| 01:18:14 | artom | My understanding is that a new microversion is the prefered approach | |
| 01:18:19 | mriedem | yes it is | |
| 01:18:44 | artom | mriedem, a spec for that seems... pulled by the hair, as we say in Quebec | |
| 01:19:01 | artom | But if folks want, I'll hack one up :) | |
| 01:19:07 | mriedem | artom: let them eat process | |
| 01:20:00 | artom | mriedem, although if we do a spec, there's no chance of it landing in ocata? | |
| 01:20:13 | mriedem | there is | |
| 01:21:09 | mriedem | artom: fwiw, | |
| 01:21:17 | mriedem | https://specs.openstack.org/openstack/nova-specs/specs/newton/implemented/response-for-invalid-status.html was for a bug fix that required a microversion in newton | |
| 01:21:26 | mriedem | it should be a relatively small and straight-forward spec | |
| 01:21:34 | mriedem | just like ^ | |
| 01:21:46 | artom | mriedem, PRECEDENT!!11 | |
| 01:23:04 | mriedem | yar | |
| 01:29:25 | edleafe | mriedem_afk: almost done with the fixes | |
| 01:56:03 | openstackgerrit | Ed Leafe proposed openstack/nova: placement: RT now adds proper Ironic inventory https://review.openstack.org/404472 | |
| 01:56:06 | edleafe | ^^ and there they are | |
| 02:03:06 | diana_clarke | I have what is probably a stupid question... when I do a `ps aux | grep nova-api` I don't see the nova-api-metadata service, but I can still successfully query the metadata service... why? | |
| 02:06:05 | diana_clarke | (asking b/c I'm looking at the # of processes started when you change osapi_compute_workers vs. metadata_workers in the nova conf) | |
| 02:08:32 | dansmith | diana_clarke: http://logs.openstack.org/30/424730/10/check/gate-grenade-dsvm-neutron-ubuntu-xenial/e3caf96/logs/etc/nova/nova.conf.txt.gz | |
| 02:08:52 | dansmith | diana_clarke: runs them both in a single process | |
| 02:09:01 | dansmith | (he said authoritatively, despite being a little fuzzy on that stuff) | |
| 02:09:30 | dansmith | diana_clarke: enabled_apis = osapi_compute,metadata | |
| 02:09:50 | openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Fix tag attribute disappearing https://review.openstack.org/426030 | |
| 02:12:42 | openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Fix tag attribute disappearing https://review.openstack.org/426030 | |
| 02:22:24 | openstackgerrit | Artom Lifshitz proposed openstack/nova: Pass APIVersionRequest to extentions https://review.openstack.org/425876 | |
| 02:22:25 | openstackgerrit | Artom Lifshitz proposed openstack/nova: Fix tag attribute disappearing in 2.33 and 2.37 https://review.openstack.org/424759 | |
| 02:23:26 | openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Fix tag attribute disappearing https://review.openstack.org/426030 | |
| 02:24:05 | diana_clarke | dansmith: sorry for quick absence (was trying to get baby to bed), scrolling back | |
| 02:24:41 | openstackgerrit | Ghanshyam Mann proposed openstack/nova: Add new policy for server list/detail with all_tenants https://review.openstack.org/415330 | |
| 02:25:56 | openstackgerrit | Ghanshyam Mann proposed openstack/nova: Add new policy for server list/detail with all_tenants https://review.openstack.org/415330 | |
| 02:34:25 | diana_clarke | dansmith: My hunch is that a new nova-api process is started for each metadata_worker, but /usr/bin/nova-api-metadata exists too (but isn't a running process). I gather it can be run independently though in some cases. I dunno... | |
| 02:35:35 | dansmith | it's used when you need them separate | |
| 02:35:41 | openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Fix tag attribute disappearing https://review.openstack.org/426030 | |
| 02:41:40 | diana_clarke | dansmith: hmmm... I think I get it now. Kind of. I'm not sure if the # of nova-api processes is osapi_compute_workers + metadata_workers + 1 or max(osapi_compute_workers, metadata_workers) + 1 | |
| 02:41:52 | diana_clarke | dansmith: and maybe maybe nproc'd, but I think I'm getting closer. Thanks! | |
| 02:42:22 | diana_clarke | oops, Jacob Two-Two | |
| 02:47:18 | openstackgerrit | Artom Lifshitz proposed openstack/nova-specs: Fix tag attribute disappearing https://review.openstack.org/426030 | |
| 02:55:37 | openstackgerrit | Ed Leafe proposed openstack/nova: placement: RT now adds proper Ironic inventory https://review.openstack.org/404472 | |
| 03:30:40 | mriedem_afk | dansmith: cells is now 21 in the gate huh | |
| 03:30:40 | mriedem | not sure there is really anything else i'm going to work on tonight | |
| 03:30:40 | mriedem | there was an old shitty steve segal movie on the other night and i'm hoping for some similar magic tonight | |
| 03:30:40 | mriedem | *steven - we're not on an informal name basis | |
| 03:31:06 | dansmith | heh | |
| 03:31:16 | dansmith | mriedem: yeah, it's moving slow | |
| 03:31:28 | dansmith | mriedem: I think this attempt on the grenade patch is going to work | |
| 03:31:35 | dansmith | just like I thought the last 12 were going to work | |
| 03:31:50 | dansmith | I'm about done for the night though | |
| 03:36:46 | mriedem | ooo a new one eh | |
| 03:39:13 | diana_clarke | mriedem: I'm sad you've worked so much from mexico ;( I submitted one patch from the airport, but then fully embraced beach life. | |
| 03:39:49 | mriedem | i can only swim so much in the same 3 pools and single pacific ocean | |
| 03:40:45 | diana_clarke | My favorite part was ditching the car for golf-cart transportation. But I was in a private city, on a private street, so it wasn;t really mexico... | |
| 03:45:52 | mriedem | edleafe: thanks for the updates, i'll review in detail in the morning | |
| 03:46:48 | mriedem | diana_clarke: i successfully ordered a cheese, and only cheese, quesadilla for my daughter using all spanish words - i made sure to gloat to my wife | |
| 03:46:53 | mriedem | high school spanish still works | |
| 03:55:22 | dansmith | mriedem: well, the race between the subnode and the scheduler must not be the problem, | |
| 03:55:53 | dansmith | because we're not having to wait at all for that to happen | |
| 03:55:54 | diana_clarke | mriedem: Nice! I really regret not having tried to learn some spanish beforehand (I took german in school). I'm now doing 10 minutes per day for when I go back. | |
| 04:00:04 | diana_clarke | (hoping to go back, I'm also not posting the pics I got of the mexican president online, surreal town was where I was at, lol) | |
| 04:03:05 | mriedem | dansmith: ok - i haven't wrapped my head around the grenade change yet and how that fixes the scheduler thing, or is meant to, but i think i get it, basically wait for as many enabled nova-compute services to show up after the upgrade as before, because the new compute service will go down and come back up so we'd wait for that | |
| 04:04:03 | mriedem | anyway i'm calling it a night | |
| 04:06:25 | diana_clarke | mriedem: If you find yourself online tomorrow, let me know if there is something I can do, so you can swim, etc. cheers | |
| 04:20:27 | gmann | alex_xu: mriedem johnthetubaguy - its ready now, can you check - https://review.openstack.org/#/c/415330/27 | |
| 07:06:35 | openstackgerrit | Pawel Koniszewski proposed openstack/nova: Mark live_migration_downtime_steps/delay as deprecated for removal https://review.openstack.org/408002 | |
| 09:28:31 | openstackgerrit | Ghanshyam Mann proposed openstack/nova: Make policy 'all_tenants_visible' a soft enforcement rule https://review.openstack.org/426128 | |
| 09:48:21 | andreykurilin | amrith: hi! | |
| 09:54:28 | johnthetubaguy | bauzas: looking at the patch of doom, eek | |
| 09:54:37 | johnthetubaguy | whats going on there :S | |
| 09:55:33 | bauzas | johnthetubaguy: in a train to Paris so probably a bit lagging | |
| 09:55:48 | bauzas | johnthetubaguy: but lemme explain the problem | |
| 09:55:50 | johnthetubaguy | bauzas: I remember you say that, safe journey | |
| 09:55:58 | johnthetubaguy | looks like we get the wrong min version? | |
| 09:56:03 | bauzas | johnthetubaguy: exactly | |
| 09:56:08 | bauzas | we get 16 as minimum | |
| 09:56:16 | bauzas | so we call the placement API | |
| 09:56:27 | johnthetubaguy | super interesting if thats broken | |
| 09:56:34 | bauzas | but we should have 15 given we have an old subnode | |