Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-01-26
10:07:11 openstackgerrit Michael Still proposed openstack/nova: Use a service account to make vendordata requests. https://review.openstack.org/415597
10:07:12 openstackgerrit Michael Still proposed openstack/nova: Optionally make dynamic vendordata failures fatal. https://review.openstack.org/416391
10:08:18 gmann johnthetubaguy: alex_xu we need to do warning inside if condition - https://review.openstack.org/#/c/415330/25/nova/policy.py
10:08:47 gmann johnthetubaguy: alex_xu i we want to warn even people did not updated file right ?
10:09:09 gmann * i think we
10:09:37 johnthetubaguy maybe...
10:09:51 gmann i mean if there is old rule overridden then warn them irrespective of updates in policy file or not
10:10:08 johnthetubaguy Oh, good point, it needs removing either way
10:10:31 gmann yea
10:10:54 johnthetubaguy I was thinking we should probably have a more general warning about folks who override a rule to it existing default value
10:11:05 johnthetubaguy but thats something for pike, as part of the policy clean up
10:13:23 gmann johnthetubaguy: but warn about what context ? 'be careful you are overriding default one' something like
10:13:36 sfinucan mikal: Didn't know you were back - was addressing johnthetubaguy's comments on https://review.openstack.org/#/q/topic:bp/vendordata-reboot-ocata
10:13:39 sfinucan As you were :)
10:14:41 johnthetubaguy sfinucan: oh you did, sweet, let me jump on that
10:14:52 sfinucan johnthetubaguy: Well, I haven't pushed them yet
10:15:17 sfinucan mikal pushed a rebase just now and I don't want to walk all over any changes he has planned
10:16:34 johnthetubaguy sfinucan: let me know
10:17:17 sfinucan BlackDex: You might be better asking that on the general #openstack channel
10:17:23 johnthetubaguy sfinucan: raj_singh has some changes up on service tokens that are very similar
10:17:52 sfinucan BlackDex: ...or waiting until tomorrow, when the feature freeze deadline has passed :)
10:17:58 johnthetubaguy had a chat to keystone folks about how the caching works, hence me spotting it
10:18:52 BlackDex sfinucan: to busy with final stuff i guese
10:19:10 sfinucan johnthetubaguy: Have a link?
10:19:11 mikal sfinucan: I've only fixed the merge conflict so far
10:19:25 mikal sfinucan: if you've got fixes for the caching and reno stuff, I encourage you to push them up
10:19:45 johnthetubaguy sfinucan: https://review.openstack.org/#/c/412634/
10:19:56 sfinucan mikal: Yeah, I've that done. Just this one I'm stuck on https://review.openstack.org/#/c/415597/4/nova/conf/vendordata.py
10:19:59 sfinucan I'll push shortly
10:20:08 mikal Ok
10:20:26 mikal I'm actually not at work today, on prescription painkillers, and in a terrible mood
10:20:35 mikal So you're much more likely to get it up in time than me
10:20:36 mikal ;)
10:23:37 openstackgerrit OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/424828
10:25:35 bauzas johnthetubaguy: sorry, I was afk this morning
10:26:03 bauzas johnthetubaguy: you probably saw the problem with the scheduler upcall patch ?
10:26:16 bauzas mriedem helped us by adding a new Depends-On
10:26:38 bauzas but I'm a bit afraid to see the patch not being merged by Ocata given all the problems we still have when upgrading
10:26:44 bauzas :'
10:26:45 bauzas :(
10:32:31 sahid johnthetubaguy: did you notice my comments here ? https://review.openstack.org/#/c/385023/
10:32:42 sahid does it make sense for you to +W it ?
10:34:07 johnthetubaguy sahid: I haven't had chance to get back to that, stuck in priority features right now
10:38:30 openstackgerrit Sam Betts proposed openstack/nova: Change order of _cleanup_deploy and _unprovision in Ironic virt https://review.openstack.org/422678
10:49:01 andreykurilin johnthetubaguy: hi! can you put final mark to the last blocker for novaclient's release https://review.openstack.org/#/c/424256/ ?)
10:59:13 Jack_V Hi folks!
10:59:22 Jack_V I'm upgrading from mitaka to newton
10:59:42 Jack_V and got this error during creation an instance from a volume
10:59:49 Jack_V InvalidInput: Invalid input received: Invalid image identifier or unable to access requested image. (HTTP 400)
11:08:53 openstackgerrit melanie witt proposed openstack/nova: Count resources to check quota for cells https://review.openstack.org/416521
11:10:18 mdbooth mfeoktistov: Did you read the ML traffic on the privsep thing?
11:10:43 mdbooth Sounds like the concensus is that you should just go ahead and use rootwrap. Oh, and do it quickly ;)
11:16:21 mfeoktistov mdbooth: Yes, thank you. I'll send patch in a minute
11:25:47 openstackgerrit Mikhail Feoktistov proposed openstack/nova: libvirt: ephemeral disk support for virtuozzo containers https://review.openstack.org/312488
11:34:56 johnthetubaguy andreykurilin: don't we support --os-auth-token and --os-token?
11:36:21 andreykurilin johnthetubaguy: I found note about --os-auth-token in the code, but there is no place if registration for that cli argument. PS: we use keystoneclient1 method for registrating identity cli arguments and it registrate just --os-token arg
11:36:34 andreykurilin *no place of
11:36:57 johnthetubaguy I guess we broke that at some point, oops
11:37:14 johnthetubaguy or we deprecated it, and only partially removed it
11:41:01 johnthetubaguy andreykurilin: are we missing OS_TOKEN environment variable reading, or does keystoneauth already do that for us?
11:45:04 andreykurilin johnthetubaguy: ok, I found commits for you:)
11:46:44 andreykurilin johnthetubaguy: os-token argument was introduced with deprecation of os-auth-token: https://github.com/openstack/python-novaclient/commit/1f11840dd84f3570330d1fcd53d1e8eea5ff7922 . Actually that commit broke env variable and the whole token auth method via cli. https://github.com/openstack/python-novaclient/commit/08c4b746b2b894a5cb178c762850a1730e1c636e made a proper deprecation for os-auth-token with warning message. And it was finnaly removed h
11:47:17 andreykurilin johnthetubaguy: So it looks like we can avoid usage of --os-auth-token, since we have a release note that it was deprecated and removed
11:47:41 johnthetubaguy yeah, that looks good, I vaguely remember deprecating those alias things, thank you!
11:48:12 andreykurilin but it sounds like you are right about env variable. keystoneauth doesn't do it for us
11:48:52 andreykurilin johnthetubaguy: should I update my patch (and lose bauzas vote) or send another one?:)
11:49:05 johnthetubaguy andreykurilin: feels like a follow on patch
11:49:16 andreykurilin ok. give me a minute
11:49:47 johnthetubaguy andreykurilin: bit worried about kevin's comment as well, feels like we should always access os_user_domain_id in a consistent way
11:49:58 johnthetubaguy but happy for that to be a follow on
11:50:56 openstackgerrit Stephen Finucane proposed openstack/nova: Cache boot time roles for vendordata. https://review.openstack.org/417780
11:50:56 openstackgerrit Stephen Finucane proposed openstack/nova: Use a service account to make vendordata requests. https://review.openstack.org/415597
11:50:57 openstackgerrit Stephen Finucane proposed openstack/nova: Optionally make dynamic vendordata failures fatal. https://review.openstack.org/416391
11:51:07 johnthetubaguy sfinucan: I see you
11:51:13 sfinucan johnthetubaguy: Cached the tokens and added renos for all patches
11:51:15 sfinucan jinx :)
11:51:19 andreykurilin johnthetubaguy: if you look at previous patch-sets(for example this one - https://review.openstack.org/#/c/424256/8/novaclient/shell.py ), I tried several ways to make that lines a bit clear, but for some reasons I did not succed in making keystone v3 work
11:51:46 johnthetubaguy andreykurilin: lets do a separate patch that adds a test for that maybe?
11:52:17 andreykurilin ok, I'll try to do it again in folowwing patch
12:00:18 openstackgerrit Andrey Kurilin proposed openstack/python-novaclient: [cli] Use environment variable for token https://review.openstack.org/425647
12:01:38 johnthetubaguy andreykurilin: I added some questions, got confused about the SecretHelper stuff
12:13:15 andreykurilin johnthetubaguy: answered:)
12:23:13 openstackgerrit Andrey Kurilin proposed openstack/python-novaclient: [cli] Use environment variable for token https://review.openstack.org/425647
12:43:26 claudiub hi folks. any idea why nova list currently shows / returns an empty list (including with --all), even if I have spawned instances?
12:55:48 openstackgerrit OpenStack Proposal Bot proposed openstack/nova: Updated from global requirements https://review.openstack.org/424828
12:56:51 sfinucan alex_xu: Regarding https://review.openstack.org/#/c/394480 - does that require any changes on the RPC version?
12:57:33 sfinucan That's object gets sent over the wire, as far as I understand, and changes to what we send there necessitate RPC bumps
12:57:37 sfinucan ...or so I thought
13:09:18 mdbooth Extreme code smell is when reviewing a 6 line change requires a code audit of 2 other modules complete with notes: https://review.openstack.org/#/c/389593/2/nova/virt/libvirt/driver.py
13:14:32 sfinucan snikitin: https://review.openstack.org/#/c/399613
13:17:00 cdent mdbooth: On some days I feel like that describes...everything
13:17:52 mdbooth cdent: Feeling all existential. I'm literally sitting here contemplating 'what does destroy_disks really *mean*'
13:18:17 cdent bauzas, mriedem, dansmith: is there somewhere for me to leap in?
13:18:39 cdent mdbooth: I think that all depends on your position on social construction of meaning
13:18:44 mdbooth lol
13:19:14 bauzas cdent: honestly, I still need to correctly understand the problem
13:19:36 cdent bauzas: Is there only one at this point?
13:20:13 bauzas cdent: yup, I fixed the functional issue
13:20:22 bauzas cdent: now, we still have the grenade problem
13:20:35 bauzas that's the blocking issue

Earlier   Later