Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-02-11
18:04:07 mriedem leakypipes must be stuck in friday limbo
18:07:18 mriedem dansmith: yeah it's definitely tight coupling but i'm not terribly worried about it right now. i wonder if setting an attribute on the instance (not persisted) would help, but idk
18:07:25 dansmith ack
18:07:37 mriedem instance.from_down_cell = True or something
18:07:53 mriedem but then you have to check if that attribute is set on the object as well
18:08:01 mriedem or we default to False i guess
18:10:06 kukacz hi, I'm getting "MigrationPreCheckError: ... CPU doesn't have compatibility." while trying to live-migrate between hosts with same model of CPU. Turned out that by mistake I got newer package with /usr/share/libvirt/cpu_map.xml differing between those 2 hosts. downgraded libvirt-bin, restarted libvirt-bin and nova-compute services but still seeing same precheck error. where does the precheck take the cpu model from?
18:11:10 kukacz (had newer package on one of the two hosts)
18:13:58 openstackgerrit Stephen Finucane proposed openstack/python-novaclient master: Microversion 2.68: Remove 'forced' live migrations, evacuations https://review.openstack.org/635131
18:14:07 sean-k-mooney mriedem_lunch: replied inline https://review.openstack.org/#/c/624842/8//COMMIT_MSG
18:14:15 stephenfin mriedem_lunch: When you're back ^
18:31:32 artom Who's our PowerVM guy? cfriesen?
18:32:19 artom I guess efried used to "it"
18:36:40 artom There's a unit test issue with to_legacy_dict() in the PowerVMLiveMigrateData... but as far as I can tell nothing actually calls to_legacy_dict anywhere
18:36:47 artom So I'm wondering if I can just get rid of it
18:40:24 artom In fact, the only place I can find *any* _legacy_dict methods is in the migrate_data.py objects and a couple of times in the libvirt migration code where we check if we got an ovo
18:40:30 cfriesen artom: I'm x86. :) used to be we called to_legacy_dict() from check_can_live_migrate_source()
18:40:34 artom I feel like that's left over from a long time ago and I can remove all of that
18:41:01 melwitt edmondsw ^
18:41:30 artom cfriesen, ah, sorry got confused about your arch allegiance ;)
18:41:47 artom You know what, I'll propose a thing that gets rid of it all, and we'll see what breaks :)
18:41:58 edmondsw artom in a mtg but will look in a few
18:43:39 artom edmondsw, cheers! I really need to relocate, but leave your reply here and I'll check eavesdrop
18:52:01 mriedem_lunch artom: i think that was dropped when we bumped the compute rpc api version to 5.0
18:52:38 mriedem https://github.com/openstack/nova/blob/stable/queens/nova/compute/rpcapi.py#L535
19:32:16 openstackgerrit Matt Riedemann proposed openstack/nova master: Move finish_resize.(start|end) notifications to helper method https://review.openstack.org/635079
19:32:16 openstackgerrit Matt Riedemann proposed openstack/nova master: Add Migration.cross_cell_move and get_by_uuid https://review.openstack.org/614012
19:32:17 openstackgerrit Matt Riedemann proposed openstack/nova master: Add Destination.allow_cross_cell_move field https://review.openstack.org/614035
19:32:17 openstackgerrit Matt Riedemann proposed openstack/nova master: Add InstanceAction/Event create() method https://review.openstack.org/614036
19:32:18 openstackgerrit Matt Riedemann proposed openstack/nova master: Change HostManager to allow scheduling to other cells https://review.openstack.org/614037
19:32:18 openstackgerrit Matt Riedemann proposed openstack/nova master: Add CrossCellWeigher https://review.openstack.org/614353
19:32:19 openstackgerrit Matt Riedemann proposed openstack/nova master: Add Instance.hidden field https://review.openstack.org/631123
19:32:19 openstackgerrit Matt Riedemann proposed openstack/nova master: Add TargetDBSetupTask https://review.openstack.org/627892
19:32:20 openstackgerrit Matt Riedemann proposed openstack/nova master: Add CrossCellMigrationTask https://review.openstack.org/631581
19:32:20 openstackgerrit Matt Riedemann proposed openstack/nova master: Execute TargetDBSetupTask https://review.openstack.org/633853
19:32:21 openstackgerrit Matt Riedemann proposed openstack/nova master: Add can_connect_volume() compute driver method https://review.openstack.org/621313
19:32:21 openstackgerrit Matt Riedemann proposed openstack/nova master: WIP: Add prep_snapshot_based_resize_at_dest compute method https://review.openstack.org/633293
19:32:22 openstackgerrit Matt Riedemann proposed openstack/nova master: WIP: Add PrepResizeAtDestTask https://review.openstack.org/627890
19:32:22 openstackgerrit Matt Riedemann proposed openstack/nova master: Add prep_snapshot_based_resize_at_source compute method https://review.openstack.org/634832
19:32:23 openstackgerrit Matt Riedemann proposed openstack/nova master: WIP: Add PrepResizeAtSourceTask https://review.openstack.org/627891
19:32:23 openstackgerrit Matt Riedemann proposed openstack/nova master: WIP: Add finish_snapshot_based_resize_at_dest compute method https://review.openstack.org/635080
19:32:24 openstackgerrit Matt Riedemann proposed openstack/nova master: Add FinishResizeAtDestTask https://review.openstack.org/635646
19:32:25 openstackgerrit Matt Riedemann proposed openstack/nova master: Execute CrossCellMigrationTask from MigrationTask https://review.openstack.org/635668
19:35:43 openstackgerrit Matt Riedemann proposed openstack/nova master: Follow up (#2) for the bw resource provider series https://review.openstack.org/634767
20:01:57 openstackgerrit Artom Lifshitz proposed openstack/nova master: Add migration param to check_can_live_migrate_destination https://review.openstack.org/634605
20:01:57 openstackgerrit Artom Lifshitz proposed openstack/nova master: [WIP] Introduce live_migration_claim() https://review.openstack.org/635669
20:01:58 openstackgerrit Artom Lifshitz proposed openstack/nova master: [WIP] Use live_migration_claim() to check dest resources https://review.openstack.org/634606
20:01:58 openstackgerrit Artom Lifshitz proposed openstack/nova master: New objects to transmit NUMA config from dest to source https://review.openstack.org/634827
20:01:59 openstackgerrit Artom Lifshitz proposed openstack/nova master: [WIP] LM: generate new NUMA XML on the dest and send it to the source https://review.openstack.org/634828
20:01:59 openstackgerrit Artom Lifshitz proposed openstack/nova master: [WIP] LM: update NUMA-related XML on the source https://review.openstack.org/635229
20:02:00 openstackgerrit Artom Lifshitz proposed openstack/nova master: Remove _legacy_dict methods https://review.openstack.org/636210
20:03:01 openstackgerrit Matt Riedemann proposed openstack/nova master: Avoid BadRequest error log on volume attachment https://review.openstack.org/581453
20:03:19 artom mriedem, ^^ The bottom-most patch there is remove the _to_dict stuff
20:03:45 artom (I assume the bot just spammed the channel with my patches, I wasn't actually connected, and eavesdrop isn't showing anything (yet?))
20:04:51 openstackgerrit Matt Riedemann proposed openstack/nova master: Provide a useful error message when trying to update non-compute services https://review.openstack.org/620667
20:12:09 artom Wait, so what's the point of default=None in Nova objects if the field doesn't actually get set to None when you init the object without it?
20:12:36 artom Is default just for "actual" values?
20:15:21 dansmith artom: it's the default for if you ask the object to set its defaults
20:15:38 dansmith remember, an object can represent something that is a fraction of the total,
20:15:40 artom dansmith, ah, so not a fallback kind of thing
20:15:48 dansmith like instance if you don't ask for all the stuff to be filled out
20:15:54 artom Fair enough.
20:16:10 dansmith if we took the defaults everywhere, we'd overwrite things in the DB with the defaults because someone didn't query for them and then called .save() to update one little field
20:16:19 artom I still have the omnipresent 'if field in obj and obj.field is not None' construct ;)
20:16:43 artom It makes sense when you explain it that way :)
20:17:02 artom s/still have/still hate/
20:17:54 artom Or obj_attr_is_set
20:18:29 artom But yeah, I see why we need both "unset" and "None", and having automatic None defaults would clobber the former with the latter
20:18:33 artom And they don't mean the same thing
20:27:46 sean-k-mooney artom: i think you can use in or hasattr with ovo too because the baseclass overlods __contains__
20:27:55 sean-k-mooney https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L321
20:28:22 sean-k-mooney so "if x in my_obj:" will work to test if the field is set
20:29:12 artom sean-k-mooney, yep. I'd hoped to avoid all that by setting a None default, but that won't be possible
20:29:29 artom sean-k-mooney, BTW, going to back to yesterday's discussion on https://review.openstack.org/#/c/634828/9/nova/virt/libvirt/driver.py@6808
20:29:45 artom When you say "cell in enumerate(instance_numa_toplogy.cells)"
20:30:05 artom That means the guest cells are *ordered* inside i_n_t, and their order represents the guest cell ID?
20:30:15 artom I mean, I know it's a list, and lists are ordered
20:30:23 artom I just didn't think we encoded anything in that order
20:30:39 sean-k-mooney artom: yes i belive so
20:30:51 artom Because the cell.id field is the *host* cell ID, right?
20:30:57 sean-k-mooney ya
20:31:15 artom OK
20:31:33 artom Hrmpf, yeah, maybe stick with building the LibvirtConfigObjects on the dest?
20:31:39 artom It seems safer :P
20:31:50 artom As a first whack at it, anyays
20:31:51 sean-k-mooney i think we can for now.
20:31:58 sean-k-mooney we really should not have too
20:32:05 sean-k-mooney but its is simpler
20:32:16 artom Yeah, I thought the same when I first tried it in Rocky
20:32:23 artom And then went to code the XML update stuff on the source
20:32:32 artom And realized i_n_t isn't enough
20:33:00 sean-k-mooney the problem is the xml generation is mix with the buisness logic
20:33:13 sean-k-mooney they should have been kept seperate but they werent
20:33:25 artom I don't disagree :S
20:33:31 artom A lot of things should have been ;)
20:33:52 sean-k-mooney so for now i think keeping the changes minimal are fro the best adn maybe we should just decuple it fully latter
20:34:33 sean-k-mooney and by later i mean as a sperate activiy in train
20:35:36 artom We'd need a quorum of virt drivers that can support NUMA LM
20:35:48 sean-k-mooney i also hate that that file is like 9000 lines long
20:35:57 sean-k-mooney not really

Earlier   Later