Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2021-05-27
17:20:41 gmann melwitt: yeah but for booting server we need real name
17:20:53 dansmith sean-k-mooney: agree, I'd rather just exclude that field
17:20:54 gmann yeah <hidden> work
17:21:09 dansmith gmann: we would need to make it take hostname or host-uuid
17:21:14 melwitt gmann: today you do but we're talking about adding the ability to use the uuid
17:21:25 gmann sean-k-mooney: but we return those field for system admin and not for project admin is difficult way to read it
17:21:41 sean-k-mooney i dont think thats hard
17:21:44 gmann melwitt: dansmith yeah i get the point of uuid
17:21:48 sean-k-mooney we now have 9 personas
17:21:51 sean-k-mooney instead of 2
17:21:52 gmann and obfuscated uuid right
17:22:09 dansmith gmann: no, can't be obfuscated
17:22:17 sean-k-mooney so we should expect that for some api system admin will see more then project admin
17:22:18 dansmith gmann: it's opaque, but not scoped or obfuscated further
17:22:20 gmann sean-k-mooney: 7 for nova
17:23:02 sean-k-mooney 7?
17:23:13 sean-k-mooney i was expecting 6 or 9
17:23:25 sean-k-mooney either domain has no menaing or it does
17:23:46 gmann actually 5. we do not have system member
17:23:55 gmann i counted the combination also
17:24:05 sean-k-mooney ideally nova itself woudl be runnign as system member but ok
17:24:33 gmann dansmith: and having mapping in DB?
17:24:52 dansmith gmann: no, no mapping. we'd have to map every project and every host
17:25:21 sean-k-mooney gmann: dansmith is suggesting just using the current uuid
17:25:32 sean-k-mooney which i think is an ok comprmise
17:25:35 dansmith gmann: using the uuid lets us look up t he host in O(1), which is what we want. It exposes a 1:1 identifier that the project admin can use, but without any details that are often embedded in hostnames
17:25:50 sean-k-mooney its slightly better then the hostname for the reason he gave above
17:26:03 sean-k-mooney but still allows effiecnt scheduling
17:26:25 gmann and with uuid non-admn or project admin cannot get host detail so make sense
17:26:34 dansmith right
17:26:48 gmann got it
17:26:59 sean-k-mooney ya we would reject a /hypervior/detail request with a project admin token
17:27:00 dansmith you're exposing a little more of the system to the project-admin, so they can direct servers, but not everything that a hostname might encode in it
17:27:12 dansmith sean-k-mooney: no, we have to support that
17:27:17 gmann btw why we do obfuscated hostID in GET API ?
17:27:19 dansmith sean-k-mooney: they want to be able to see some details of the hosts
17:27:20 sean-k-mooney well ok yes
17:27:23 sean-k-mooney but not all the fileds
17:27:33 dansmith gmann: so that users can't tell that they are on the same host across coke/pepsi
17:27:45 gmann ohk,
17:28:16 openstackgerrit Lee Yarwood proposed openstack/nova master: WIP libvirt: Do not destroy volume secrets during _hard_reboot https://review.opendev.org/c/openstack/nova/+/793463
17:28:29 dansmith gmann: okay, so .. all good?
17:29:02 gmann one last one, on new policy
17:29:04 sean-k-mooney there is one other advantage
17:29:22 gmann this one https://review.opendev.org/c/openstack/nova-specs/+/793011/2/specs/xena/approved/allow-project-admin-list-hypervisors.rst#41
17:29:24 sean-k-mooney if im not mistaken we do not support using the uuid with the old az hack
17:29:49 sean-k-mooney so they can bypass the sculder with --availabit-zone <zone>:<hostname>
17:29:49 gmann any objection on adding new policy to let operator to list all hypervisors ?
17:30:00 sean-k-mooney if we give the the uuid right
17:30:31 dansmith gmann: no, I think that makes sense if they want to allow that
17:30:58 gmann cool
17:31:04 gmann I will update the spec accordingly.
17:31:09 gmann thanks dansmith melwitt sean-k-mooney
17:31:19 dansmith gmann: cool, sorry for missing your reply earlier
17:31:22 sean-k-mooney oh to remove the aggreate metadtata requirement
17:31:37 sean-k-mooney form an end user perspecitv i would not be able to tell the diffrenece
17:31:38 gmann dansmith: np! I too missed to ask that day itself.
17:32:04 sean-k-mooney so that policy would not be an interop issue so i think its ok to add one yes
17:33:05 sean-k-mooney ok i am going to finish up for the day. ill see everyon on our new home on tuesday o/
17:33:25 dansmith sean-k-mooney: enjoy o/
17:33:26 melwitt have a good weekend o/
17:35:21 dansmith wow
17:35:28 dansmith not used to seeing that :P
17:36:17 melwitt 😆
17:39:21 artom He's been making big improvement in keeping a regular work schedule
17:39:26 artom (I sound like his dad)
17:39:54 gmann lucky redhatters :) long weekend
17:40:24 gmann I miss Tokyo in that term for many long weekend and three golden weeks in a year
23:23:38 openstackgerrit Ghanshyam proposed openstack/nova-specs master: Allow project admin to list hypervisors https://review.opendev.org/c/openstack/nova-specs/+/793011
23:23:52 gmann dansmith: melwitt ^^
#openstack-nova - 2021-05-28
00:42:54 openstackgerrit melanie witt proposed openstack/nova master: Honor [neutron]http_retries in the manual client https://review.opendev.org/c/openstack/nova/+/793512
01:34:47 openstackgerrit Wenping Song proposed openstack/nova master: Replaces tenant_id with project_id from Rebuild Server API https://review.opendev.org/c/openstack/nova/+/766380
01:38:59 openstackgerrit Wenping Song proposed openstack/nova master: Replaces tenant_id with project_id from List/Update Servers APIs https://review.opendev.org/c/openstack/nova/+/764292
08:39:04 openstackgerrit Wenping Song proposed openstack/nova master: Replace tenant* with project* in codes https://review.opendev.org/c/openstack/nova/+/769329
08:46:38 yonglihe sean-k-mooney, gibi, do you have time for smart patches? thanks. https://review.opendev.org/q/topic:%22bp%252Fsriov-smartnic-support%22+(status:open%20OR%20status:merged)
09:30:13 openstackgerrit Merged openstack/nova-specs master: [pps] Move the alternatives to separate section https://review.opendev.org/c/openstack/nova-specs/+/792634
11:45:48 openstackgerrit Balazs Gibizer proposed openstack/nova-specs master: Allow provider re-parenting in placement https://review.opendev.org/c/openstack/nova-specs/+/788243
11:52:31 openstackgerrit Balazs Gibizer proposed openstack/nova master: Add test coverage for security group checking https://review.opendev.org/c/openstack/nova/+/786044
12:15:43 openstackgerrit Balazs Gibizer proposed openstack/nova-specs master: [template]suggest work item ordering https://review.opendev.org/c/openstack/nova-specs/+/793197
12:38:00 openstackgerrit Balazs Gibizer proposed openstack/nova master: [func test] create pps resource on OVS agent RP https://review.opendev.org/c/openstack/nova/+/787205
12:38:00 openstackgerrit Balazs Gibizer proposed openstack/nova master: Fix RequestLevelParams persistence handling in RequestSpec https://review.opendev.org/c/openstack/nova/+/791502
12:38:01 openstackgerrit Balazs Gibizer proposed openstack/nova master: [func test] refactor assertPortMatchesAllocation https://review.opendev.org/c/openstack/nova/+/792458
12:38:01 openstackgerrit Balazs Gibizer proposed openstack/nova master: [func test] move port creation to the NeutronFixture https://review.opendev.org/c/openstack/nova/+/787206
12:40:38 elod gibi: fyi, there are release patches to review:
12:40:40 elod https://review.opendev.org/c/openstack/releases/+/792924
12:40:40 elod https://review.opendev.org/c/openstack/releases/+/792923
12:40:41 elod https://review.opendev.org/c/openstack/releases/+/792938
12:41:04 elod for me they seem OK :]
12:43:00 openstackgerrit Balazs Gibizer proposed openstack/nova master: Add same_subtree field to RequestLevelParams https://review.opendev.org/c/openstack/nova/+/791503
12:45:05 gibi elod: ack
12:46:19 openstackgerrit Balazs Gibizer proposed openstack/nova master: Bump min placement microversion to 1.36 https://review.opendev.org/c/openstack/nova/+/791504
12:48:01 openstackgerrit Balazs Gibizer proposed openstack/nova master: Support same_subtree in allocation_canadidate query https://review.opendev.org/c/openstack/nova/+/791505
12:51:38 gibi elod: the osc-placement patch needs more cores as it needs a major version bump. Can we get back to that then RH cores are back next week?
12:53:08 gibi s/then/when/
12:55:47 elod gibi: I think yes
12:55:58 elod though yesterday was the M1 deadline
12:56:14 gibi elod: cool. I'm not confident bumping the major version alone while they are out
12:56:39 elod gibi: ++
16:24:20 openstackgerrit Balazs Gibizer proposed openstack/nova master: Detect extended_resource_request neutron API extension https://review.opendev.org/c/openstack/nova/+/793618
16:25:49 openstackgerrit Balazs Gibizer proposed openstack/nova master: Support the new port resource_request format https://review.opendev.org/c/openstack/nova/+/787208

Earlier   Later