| *** openstack has joined #openstack-placement | 17:05 | |
| *** ChanServ sets mode: +o openstack | 17:05 | |
| openstackgerrit | Lee Yarwood proposed openstack/nova master: libvirt: Report the virtual size of RAW disks https://review.openstack.org/567899 | 17:11 |
|---|---|---|
| *** mriedem_lunch is now known as mriedem | 17:40 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: DNM: test raw image live migration with vfat config drive https://review.openstack.org/567860 | 17:45 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: add consumers generation field https://review.openstack.org/557958 | 17:51 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: placement: Allocation.consumer field https://review.openstack.org/565405 | 17:51 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: rework allocation handler _allocations_dict() https://review.openstack.org/565407 | 17:51 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: Add a microversion for consumer generation support https://review.openstack.org/565604 | 17:51 |
| leakypipes | superdan, cdent, fried_rolls: ^^ rebased the consumer generation series. removed all the stuff about consumer/project/user cleanups (that is all in 557958 now). | 17:52 |
| superdan | leakypipes: did you see I shat upon your bottom patch? | 17:52 |
| leakypipes | superdan: I did. I'll address those shortly. | 17:53 |
| superdan | ack | 17:53 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/nova master: Fix HTTP500 error of changes-since on v2.0 API https://review.openstack.org/567916 | 17:56 |
| openstackgerrit | Chris Dent proposed openstack/nova master: Flexibly test keystonmiddleware in placement stack https://review.openstack.org/567923 | 18:18 |
| openstackgerrit | Merged openstack/nova master: libvirt: always pass emulator threads policy https://review.openstack.org/561913 | 18:36 |
| openstackgerrit | Merged openstack/nova master: Add connection_parameters to list of items copied from database https://review.openstack.org/567248 | 18:39 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Don't reschedule on RequestedVRamTooHigh errors https://review.openstack.org/567929 | 18:47 |
| *** mriedem is now known as mriedem_afk | 18:57 | |
| *** fried_rolls is now known as fried_rice | 19:03 | |
| fried_rice | #startmeeting how_bout_now | 19:06 |
| openstack | Meeting started Fri May 11 19:06:31 2018 UTC and is due to finish in 60 minutes. The chair is fried_rice. Information about MeetBot at http://wiki.debian.org/MeetBot. | 19:06 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 19:06 |
| *** openstack changes topic to " (Meeting topic: how_bout_now)" | 19:06 | |
| openstack | The meeting name has been set to 'how_bout_now' | 19:06 |
| fried_rice | woot | 19:06 |
| fried_rice | https://bugs.launchpad.net/nova/+bug/1770570 | 19:06 |
| openstack | Launchpad bug 1770570 in OpenStack Compute (nova) "In a parent/child resource provider aggregate, if a child resource providers is deleted then the deleted child resource provider will be seen as available in the parent resource provider aggregate list." [Undecided,New] | 19:06 |
| fried_rice | #endmeeting | 19:06 |
| *** openstack changes topic to "See https://docs.openstack.org/nova/latest/user/placement.html and https://developer.openstack.org/api-ref/placement/" | 19:06 | |
| openstack | Meeting ended Fri May 11 19:06:55 2018 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 19:06 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/how_bout_now/2018/how_bout_now.2018-05-11-19.06.html | 19:06 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/how_bout_now/2018/how_bout_now.2018-05-11-19.06.txt | 19:07 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/how_bout_now/2018/how_bout_now.2018-05-11-19.06.log.html | 19:07 |
| fried_rice | and http://eavesdrop.openstack.org/irclogs/%23openstack-placement/%23openstack-placement.2018-05-11.log.html since about 2h ago. | 19:07 |
| fried_rice | cdent: ^^ | 19:07 |
| cdent | woot | 19:08 |
| figleaf | fancy! | 19:09 |
| fried_rice | Who's gonna run the sched meeting on Monday? cdent figleaf | 19:14 |
| fried_rice | leakypipes: | 19:14 |
| fried_rice | gibi: alex_xu | 19:15 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: placement: always create consumer records https://review.openstack.org/567678 | 19:15 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: add consumers generation field https://review.openstack.org/557958 | 19:15 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: placement: Allocation.consumer field https://review.openstack.org/565405 | 19:15 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: rework allocation handler _allocations_dict() https://review.openstack.org/565407 | 19:15 |
| openstackgerrit | Jay Pipes proposed openstack/nova master: Add a microversion for consumer generation support https://review.openstack.org/565604 | 19:15 |
| leakypipes | fried_rice: I'll be around. | 19:15 |
| fried_rice | cool | 19:16 |
| fried_rice | cdent: So is it either keystone_uri='foo' or keystone_uri="foo" now? | 19:24 |
| cdent | yeah, '' in <= 5.0.0 "" in 5.1.x | 19:24 |
| fried_rice | cdent: okay, left a comment. I'm sure it's okay as is, but possible to regress if auth_uri is a substring of what you expect. | 19:31 |
| fried_rice | cdent: And I thought keystone guys were always really careful about compat changes like that. | 19:31 |
| cdent | keystone != keystonemiddleware | 19:33 |
| cdent | I responded with a comment on why i prefer what it's doing | 19:34 |
| fried_rice | leakypipes, mriedem_afk: Since we went through weeks of mud wrestling to get to where group_policy is required to the placement API... should it be required similarly in the flavor, or should we default? | 20:09 |
| *** mriedem_afk is now known as mriedem | 20:26 | |
| mriedem | that's only if you're using granular right? | 20:26 |
| mriedem | and how would you enforce it? when extra specs are added to the flavor? | 20:26 |
| leakypipes | fried_rice: if >1 request group results from the flavor extra spec, then yes, I think we should require a group_policy extra spec. | 20:28 |
| leakypipes | though mriedem does have a good point about when you'd enforce it. | 20:28 |
| leakypipes | tricky. | 20:28 |
| fried_rice | mriedem: Enforcement will happen (whether we like it or not) at the placement API regardless. | 20:28 |
| fried_rice | I can add extra enforcement when parsing the flavor. | 20:29 |
| fried_rice | Do we do enforcement type stuff at flavor creation generally speaking? | 20:29 |
| fried_rice | leakypipes: We do *some* of that kind of enforcement when we parse the flavor - like "did you specify a bogus resource class name", but not "do you have a requiredN without a corresponding resourcesN". How much of that do we want to have? | 20:30 |
| fried_rice | effectively duplicating logic from the placement API schema / handler / code. | 20:31 |
| mriedem | when parsing the flavor, you mean during the server create / scheduling request? | 20:34 |
| mriedem | in that case you'd fail the user request for a thing the admin f'ed up | 20:34 |
| mriedem | hopefully admins/operators are testing flavors before they make them public, but... | 20:34 |
| mriedem | we've never done flavor extra spec validation at the time of creating extra specs though, so idk | 20:35 |
| mriedem | it doesn't seem crazy to me though | 20:35 |
| mriedem | superdan might think it's crazy | 20:35 |
| mriedem | CRAZY HOT! | 20:35 |
| fried_rice | Okay, so at the moment in this patch I'm just working on the parsing during scheduling, and I can add a check in there to make sure group_policy is specified if >1 granular group; that's easy enough. If we want to go add some parsing/checking/enforcement at flavor creation later on, we can do that too. | 20:37 |
| openstackgerrit | Merged openstack/nova master: Fix irrelevant-files in nova-dsvm-multinode-base https://review.openstack.org/567884 | 20:53 |
| leakypipes | fried_rice: not sure how we'd do that, considering someone can add flavor extra specs one at a time... | 20:53 |
| fried_rice | leakypipes: mm. | 20:53 |
| fried_rice | mriedem: I think there's your answer ^ Gonna keep it at sched time for the forseeable. | 20:54 |
| mriedem | fried_rice: i'm ok with at scheduling time for now | 20:56 |
| fried_rice | ack | 20:56 |
| mriedem | leakypipes: yup i thought about that as well, | 20:56 |
| mriedem | so you could either add the policy while adding the granular specs in the same POST/PUT flavor extra specs call, or we'd return a 400 if you tried adding granular without policy (or not already on the flavor) | 20:57 |
| mriedem | but yeah, it's new territory | 20:57 |
| mriedem | ultimately, operators should test the flavors they are creating before making them public, | 20:57 |
| mriedem | ESPECIALLY if they have granular request groups in them because those are crazy | 20:57 |
| mriedem | and likely easy to fudge up | 20:57 |
| mriedem | POST /flavors/{flavor_id}/validate :) | 20:58 |
| mriedem | POST /flavors/{flavor_id}/dryhump | 20:58 |
| *** avolkov has quit IRC | 20:59 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: DNM: test raw image live migration with vfat config drive https://review.openstack.org/567860 | 20:59 |
| fried_rice | mriedem: From a CLI perspective I was thinking it could do validation without enforcement, and issue warnings if discrepancies were found. That way you would be able to issue subsequent commands to remedy warnings as you see them. But not sure how that works going through the API. | 21:02 |
| cdent | fried_rice: on that regex thing: no, we don't care: we're not testing keystonemiddleware | 21:03 |
| fried_rice | okay. | 21:03 |
| mriedem | fried_rice: but which CLI? | 21:05 |
| mriedem | pick your poison | 21:05 |
| fried_rice | vay | 21:05 |
| mriedem | trick, they are all poison | 21:05 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: WIP: xenapi: drop deprecated vif_driver config option https://review.openstack.org/567877 | 21:15 |
| openstackgerrit | Eric Fried proposed openstack/nova master: Granular requests to get_allocation_candidates https://review.openstack.org/515811 | 21:50 |
| fried_rice | mriedem: Can you lay hands on a reno I can use as an example for https://review.openstack.org/#/c/515811/12 -- i.e. one for where we've changed/added extra_specs syntax (fairly dramatically in this case)? | 21:51 |
| fried_rice | cdent: Did you add a reno for forbidden traits in extra specs? | 21:53 |
| fried_rice | aha, got it. cdent mriedem cancel, thanks. | 21:54 |
| *** figleaf is now known as edleafe | 21:54 | |
| *** edmondsw has quit IRC | 21:58 | |
| *** cdent has quit IRC | 22:00 | |
| openstackgerrit | Merged openstack/nova master: Address feedback from instance_list smart-cell behavior https://review.openstack.org/567643 | 22:00 |
| openstackgerrit | Merged openstack/nova stable/pike: libvirt: check image type before removing snapshots in _cleanup_resize https://review.openstack.org/567625 | 22:00 |
| *** edmondsw has joined #openstack-placement | 22:02 | |
| *** edmondsw has quit IRC | 22:06 | |
| openstackgerrit | Eric Fried proposed openstack/nova master: Granular requests to get_allocation_candidates https://review.openstack.org/515811 | 22:16 |
| fried_rice | gibi: leakypipes: blayum, I think this is ready for initial review ^ | 22:16 |
| fried_rice | aaaand I'm out. | 22:16 |
| *** fried_rice is now known as efried | 22:16 | |
| *** edmondsw has joined #openstack-placement | 22:29 | |
| openstackgerrit | Merged openstack/nova master: update scheduler to use image-traits https://review.openstack.org/560596 | 22:34 |
| openstackgerrit | Tsuyoshi Nagata proposed openstack/nova master: nova improvement of maximum attach volumes more than 26 vols https://review.openstack.org/567472 | 23:15 |
| openstackgerrit | Eric Fried proposed openstack/nova master: Granular requests to get_allocation_candidates https://review.openstack.org/515811 | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!