Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-02-22
22:31:32 sean-k-mooney melwitt: https://github.com/openstack/nova/commit/6e49019fae80586c4bbb8a7281600cf6140c176a
22:31:46 melwitt request spec has a wild west reputation, so I'm just not sure if I could rely on its user_id field. being that I didn't dig into it yet
22:32:18 melwitt ah, wow. relatively new
22:32:37 sean-k-mooney well its been there since queens?
22:32:52 melwitt and no data migration, so would be null for already existing request specs when that code landed
22:33:19 melwitt yeah, I just expected it would be older than that
22:33:30 melwitt ok, nevermind my request spec flighty idea
22:34:35 sean-k-mooney well looking at the change it look like there are check in the conductor that depend on both the user id and porject id being set
22:34:56 mriedem fried_rice: well, i got something working for the functional regression test at least
22:35:00 melwitt hah, ok
22:35:05 mriedem of course it's rife with TODOs and FIXMEs
22:35:09 fried_rice neat.
22:35:17 sean-k-mooney i wasnt following hte context of what you were working on but it look like it will alway be none none
22:36:52 sean-k-mooney melwitt: you could also get the user_id out of the instance if you have that https://github.com/openstack/nova/blob/master/nova/objects/instance.py#L122-L123
22:40:04 melwitt sean-k-mooney: I'm adding user_id to InstanceMapping and need a reliable way to populate it (data migration of already existing records). and instance was the assumed way. and it occurred to me (from looking at the down cell stuff) that, request spec has user_id too, could I use that? the answer I now know is, no
22:40:26 melwitt I already wrote everything to use instance user_id, so this means no change for me
22:40:50 sean-k-mooney ah ok
22:44:00 sean-k-mooney melwitt: you could be checky and look at the allocation for the instacne and get the user id out of the consumers table in the api db
22:44:44 sean-k-mooney the consumer and allocation should have been there since pike
22:51:29 sean-k-mooney melwitt: :) oh look like you added the consumer table in pike https://github.com/openstack/nova/commit/b9eed6a2862d0d72c77706b49d05ea588d2d81f3
23:04:08 melwitt yeah, I dunno
23:06:00 melwitt that might not work bc I need to migrate soft-deleted instances (the soft-delete API) too, since they could restored theoretically at any point in the future. and I don't think those would have allocations, as they are supposed to be gone from the user perspective and not consuming resources
23:07:29 openstackgerrit Merged openstack/nova stable/pike: Not set instance to ERROR if set_admin_password failed https://review.openstack.org/608180
23:07:31 melwitt besides that, it seems most reliable to use instances. we had bugs in the past with orphaned allocations in placement and deployments out there might still have some
23:07:36 openstackgerrit Merged openstack/nova stable/pike: Add functional regression test for bug 1806064 https://review.openstack.org/623935
23:07:37 sean-k-mooney ya i guess you have to blance the risks. if the cell is donw then using ht user form the instance will fail
23:07:37 openstack bug 1806064 in OpenStack Compute (nova) pike "Volume remains in attaching/reserved status, if the instance is deleted after TooManyInstances exception in nova-conductor" [Medium,In progress] https://launchpad.net/bugs/1806064 - Assigned to Matt Riedemann (mriedem)
23:08:24 sean-k-mooney melwitt: ya the instace is proably the safest whcih is why i said using the consumer recored associated with the instance allocation was a little cheaky
23:09:15 melwitt oh, cheaky? you spelled it "checky" earlier and I didn't get it :)
23:09:37 sean-k-mooney oh ha ya
23:10:24 sean-k-mooney well it shoudl be cheeky if i use the irish/uk spelling
23:10:52 sean-k-mooney anyway i better go. have a good weekend o/
23:11:04 melwitt have a good weekend o/
23:21:28 openstackgerrit Merged openstack/nova stable/pike: Create BDMs/tags in cell with instance when over-quota https://review.openstack.org/623937
23:22:40 mriedem hmm, why do we persist RequestSpec.ignore_hosts
23:22:46 mriedem dansmith: ^ seems dangerous yeah?
23:22:55 mriedem just like why we don't persist retry and requested_destination
23:23:41 mriedem i.e. resize with allow_resize_to_same_host=false sets ignore_hosts=[instance.host], we save that, then resize again with allow_resize_to_same_host=False will always filter out the same host
23:24:02 mriedem er allow_resize_to_same_host=True the 2nd run
23:24:28 mriedem i know request spec is the thing everyone wants to think about at 3:30 on a friday
23:24:32 sean-k-mooney do we clear it on resize confirm/revert
23:24:55 sean-k-mooney i dont recal if we can retry a resize but perhapse that is why
23:25:09 sean-k-mooney that is just a guess
23:25:18 mriedem well, on the 2nd resize we would overwrite RequestSpec.ignore_hosts again
23:26:08 mriedem looks like the only other move operation that really looks at it is evacuate, and you can't evacuate to the same host anyway so ..
23:26:24 mriedem evacuate ignores instance.host as well
23:26:35 mriedem so i guess this is why it's not a problem
23:27:27 mriedem live migrate also uses it while retrying different hosts in conductor, but doesn't persist those changes (until someone comes along and adds code to the live migration task that saves the request spec for some reason)
23:29:51 sean-k-mooney well it doesnt sound like there is an obvious reason to save it to the db so maybe put up a patch to and see what breaks but form the sounds of it might not be needed and might be a source of a future bug
23:30:26 mriedem no time right now, but yeah i'm sure it would be a future bug, i.e. https://review.openstack.org/#/c/636271/
23:30:39 mriedem these per-operation fields on request spec keep biting us in our collective asses
23:43:53 openstackgerrit Matt Riedemann proposed openstack/nova master: WIP: Workaround resize to same host doubling allocations bug https://review.openstack.org/638791
23:44:05 mriedem leakypipes: fried_rice: ^ damn near killed me
23:44:07 mriedem fugly as hell
23:45:46 openstackgerrit sean mooney proposed openstack/nova master: [DNM] stop storing request_spec.ignore_hosts in the db. https://review.openstack.org/638792
23:46:13 sean-k-mooney i belive ^ is syantaticlly corret python. beyond that its just a reminder
23:46:37 sean-k-mooney but we can see if any of the ci job fall over becase of it
23:47:51 sean-k-mooney ok im actully leaving this time o/
#openstack-nova - 2019-02-23
00:31:42 openstackgerrit Merged openstack/nova master: FUP for Id7827fe8dc27112e342dc25c902c8dbc25f63b94 https://review.openstack.org/638370
00:43:29 openstackgerrit Merged openstack/nova master: vmware:add support for the hw_video_ram image property https://review.openstack.org/564193
00:43:37 openstackgerrit Merged openstack/nova master: Remove _legacy_dict methods https://review.openstack.org/636210
00:43:45 openstackgerrit Merged openstack/nova stable/queens: PCI: do not force remove allocated devices https://review.openstack.org/635072
00:43:53 openstackgerrit Merged openstack/nova stable/rocky: Fix typo https://review.openstack.org/633196
00:44:04 openstackgerrit Merged openstack/nova master: Fix irrelevant-files for legacy-grenade-dsvm-neutron-multinode-live-migration https://review.openstack.org/637231
06:10:46 openstackgerrit Merged openstack/nova master: Remove deprecated 'os-server-groups' policy https://review.openstack.org/633672
06:13:05 openstackgerrit Merged openstack/nova master: Remove deprecated 'flavors' policy https://review.openstack.org/633664
06:33:08 openstackgerrit Merged openstack/nova master: Fix race in test_volume_swap_server_with_error https://review.openstack.org/634334
09:22:33 openstackgerrit Merged openstack/nova master: libvirt: Rewrite _create_pty_device() to be clearer https://review.openstack.org/637152
09:22:42 openstackgerrit Merged openstack/nova master: libvirt: Drop MIN_LIBVIRT_PARALLELS_SET_ADMIN_PASSWD https://review.openstack.org/632514
09:44:16 openstackgerrit Zhenyu Zheng proposed openstack/nova master: Bump compute service to indicate attach/detach root volume is supported https://review.openstack.org/614750
09:56:54 openstackgerrit Zhenyu Zheng proposed openstack/nova master: WIP detach root volume API changes https://review.openstack.org/623981
10:13:47 openstackgerrit Vlad Gusev proposed openstack/nova stable/queens: Provide a useful error message when trying to update non-compute services https://review.openstack.org/638803
12:18:46 openstackgerrit Takashi NATSUME proposed openstack/nova master: Override the 'get' method in DriverBlockDevice class https://review.openstack.org/638821
12:20:00 openstackgerrit Takashi NATSUME proposed openstack/nova master: Override the 'get' method in DriverBlockDevice class https://review.openstack.org/638821
13:12:34 openstackgerrit Merged openstack/nova stable/queens: Handle unicode characters in migration params https://review.openstack.org/636678
14:02:34 openstackgerrit Balazs Gibizer proposed openstack/nova master: Use Selection object to fill request group mapping https://review.openstack.org/638711
14:22:48 openstackgerrit Balazs Gibizer proposed openstack/nova master: Add trait cache for multicreate https://review.openstack.org/638847
14:58:55 openstackgerrit Merged openstack/nova master: Fix a typo in configuration description https://review.openstack.org/638445
14:59:04 openstackgerrit Merged openstack/nova master: Refactor bdm handling in ServersController.create method https://review.openstack.org/633535
14:59:12 openstackgerrit Merged openstack/nova master: Move finish_resize.(start|end) notifications to helper method https://review.openstack.org/635079
15:35:50 openstackgerrit Takashi NATSUME proposed openstack/python-novaclient master: Fix changes-before values in an instance action test https://review.openstack.org/638849
20:06:56 openstackgerrit Matt Riedemann proposed openstack/nova master: Refactor "networks" processing in ServersController.create https://review.openstack.org/633594
20:36:40 openstackgerrit Merged openstack/nova master: Cleanup return_reservation_id in ServersController.create https://review.openstack.org/633560
21:44:42 openstackgerrit Artom Lifshitz proposed openstack/nova master: Introduce live_migration_claim() https://review.openstack.org/635669
21:44:42 openstackgerrit Artom Lifshitz proposed openstack/nova master: Add migration param to check_can_live_migrate_destination https://review.openstack.org/634605
21:44:43 openstackgerrit Artom Lifshitz proposed openstack/nova master: LM: add support for sending NUMAMigrateData to the source https://review.openstack.org/634828
21:44:43 openstackgerrit Artom Lifshitz proposed openstack/nova master: New objects to transmit NUMA config from dest to source https://review.openstack.org/634827
21:44:44 openstackgerrit Artom Lifshitz proposed openstack/nova master: [WIP needs more tests] Full NUMA live migration support https://review.openstack.org/634606
21:44:44 openstackgerrit Artom Lifshitz proposed openstack/nova master: LM: add support for updating NUMA-related XML on the source https://review.openstack.org/635229
21:44:45 openstackgerrit Artom Lifshitz proposed openstack/nova master: Revert "Fail to live migration if instance has a NUMA topology" https://review.openstack.org/638868
21:55:22 openstackgerrit Artom Lifshitz proposed openstack/nova master: Add migration param to check_can_live_migrate_destination https://review.openstack.org/634605
21:55:23 openstackgerrit Artom Lifshitz proposed openstack/nova master: New objects to transmit NUMA config from dest to source https://review.openstack.org/634827
21:55:23 openstackgerrit Artom Lifshitz proposed openstack/nova master: Introduce live_migration_claim() https://review.openstack.org/635669
21:55:24 openstackgerrit Artom Lifshitz proposed openstack/nova master: LM: add support for updating NUMA-related XML on the source https://review.openstack.org/635229
21:55:24 openstackgerrit Artom Lifshitz proposed openstack/nova master: LM: add support for sending NUMAMigrateData to the source https://review.openstack.org/634828
21:55:25 openstackgerrit Artom Lifshitz proposed openstack/nova master: Revert "Fail to live migration if instance has a NUMA topology" https://review.openstack.org/638868
21:55:25 openstackgerrit Artom Lifshitz proposed openstack/nova master: [WIP needs more tests] Full NUMA live migration support https://review.openstack.org/634606
#openstack-nova - 2019-02-24
00:13:48 openstackgerrit Merged openstack/nova master: Log why rescheduling is disabled https://review.openstack.org/638699
00:27:00 openstackgerrit Merged openstack/nova master: Remove misleading code from _move_operation_alloc_request() https://review.openstack.org/638700
01:19:57 openstackgerrit Zhenyu Zheng proposed openstack/nova master: Bump compute service to indicate attach/detach root volume is supported https://review.openstack.org/614750

Earlier   Later