Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-09-25
15:36:56 dansmith mriedem: just think of all the awesome we get once that is in:)
15:37:14 dansmith -353, +91
15:37:17 dansmith pretty good ratio
15:38:30 mriedem if you like that
15:39:16 mriedem oh nvm
15:39:21 mriedem it already merged https://review.openstack.org/#/c/460377/ :)
15:39:54 mriedem stephenfin: an excellent +1 +W there ^
15:40:22 dansmith oh, I thought you mean the patch that uses the instance_list bit
15:40:33 stephenfin mriedem: God damn it
15:40:59 stephenfin dansmith: Fancy taking a look through it anyway? If I missed something, I imagine we can revert
15:41:39 dansmith stephenfin: in a bit, I'm in the middle of rebase hell with my migration uuid set
15:41:57 stephenfin (y)
15:52:31 openstackgerrit Jianghua Wang proposed openstack/nova-specs master: Support virtual GPU resources https://review.openstack.org/450122
15:54:22 openstackgerrit Balazs Gibizer proposed openstack/nova master: cover migration cases with functional tests https://review.openstack.org/493865
15:54:39 mriedem ooo pike 16.0.1 was released
15:55:05 bauzas jaypipes: mriedem: dansmith: as we discussed previously, if we say we could just imagine implementing baby steps for VGPU tracking by just amending the current RP that is returned by the virt driver and just add some resource class about how many VGPUs are available, that would mean it would be limited to only one single track of the virtual GPUs, whatever the number of devices we have
15:55:33 jaypipes bauzas: yeah
15:55:39 dansmith bauzas: which would be fine I think
15:55:40 bauzas jaypipes: mriedem: dansmith: to make it clear, once nested RPs would be a thing, we would have the virt driver reporting the inventory per pGPU
15:55:46 dansmith yes
15:55:53 jaypipes bauzas: until n-r-p there would only be a single provider to associate a trait to, so yeah
15:55:59 bauzas okay, we're all in violent agreement then
15:56:07 jaypipes VIOLENCE!
15:56:18 mriedem bauzas: are you going to implement get_inventory for the virt driver then? or munge it into get_available_resources?
15:56:29 mriedem i think only ironic and libvirt have the get_inventory method implemented
15:56:30 bauzas mriedem: I was planning to use the new call
15:56:33 mriedem ok
15:56:34 dansmith yes
15:56:41 mriedem jianghuaw would do it for xen,
15:56:54 bauzas mriedem: yeah, libvirt already uses the new interface
15:56:54 mriedem and either cdent or rado (sp?) is i think doing it for vmware
15:57:06 mriedem radu?
15:57:27 mriedem and according to claudiub, there is nothing hyperv *can't* do, so i'm sure he's up for implementing that
15:57:33 jaypipes heh
15:57:50 bauzas HAH
15:58:17 bauzas honestly, I just feel having vGPU support in Nova would be an interesting usecase for showing how Placement can help :)
15:58:35 bauzas and no longer do any crap stuff like PCIDeviceTracker
15:58:43 jianghuaw mriedem, absolutely I'm happy to take the needed work for xen.
15:58:54 mriedem cdent: edleafe: one of you want to update this for nova? http://specs.openstack.org/openstack/api-wg/liaisons.html#liaisons
15:59:18 jaypipes lyarwood: I can't believe you didn't list my proposed name of "ffs" for fast-forward skips ;)
15:59:44 dansmith jaypipes: because skip can't be in the name
15:59:54 edleafe mriedem: who should the new liaison be?
15:59:56 dansmith but if we could come up with another word for the s
15:59:57 jaypipes dansmith: I know I was only kidding. :)
16:00:18 mriedem how about slop
16:00:26 mriedem skip-level offline (upgrade) process
16:00:34 jianghuaw bauzas, jaypipes: please help to check if you are happy with the revised vGPU spec: https://review.openstack.org/#/c/450122/
16:00:38 dansmith can't have skip in the name
16:00:43 jaypipes jianghuaw: yep, will do.
16:00:46 bauzas jianghuaw: I'm already on it
16:01:15 claudiub what do I have to do?
16:01:34 jaypipes dansmith: I still like the package repository - integrated clustered kubernetes
16:01:35 mriedem claudiub: make the hyperv driver implement the get_inventory method
16:01:52 mriedem claudiub: to set the stage for vgpu support https://review.openstack.org/#/c/450122/
16:02:11 claudiub awesome
16:03:25 jianghuaw jaypipes, bauzas: thanks both:-)
16:04:37 claudiub welp, at the moment hyper-v reports some vgpu resources / stats, so i guess we'll have to move that around a bit.
16:10:08 jianghuaw bauzas, cool. thanks for +2 on the spec:-)
16:11:45 openstackgerrit Dan Smith proposed openstack/nova master: Use improved instance_list module in compute API https://review.openstack.org/505418
16:11:46 openstackgerrit Dan Smith proposed openstack/nova master: Remove legacy fault-loading routines https://review.openstack.org/505456
16:11:46 openstackgerrit Dan Smith proposed openstack/nova master: Fix CellDatabases fixture swallowing exceptions https://review.openstack.org/506312
16:11:47 openstackgerrit Dan Smith proposed openstack/nova master: Fix minor input items from previous patches https://review.openstack.org/506416
16:13:23 dansmith mriedem: I just pushed ^ so that that functional change isn't in the series
16:13:50 mriedem are you going to abandon that one?
16:13:59 dansmith sure
16:14:33 yufei could someone help review this, a small patch which remove re-auth logic from ironic client wrapper https://review.openstack.org/#/c/502382/
16:15:22 yufei keystone session supports to re-auth for 401 error, so re-auth in ironic client is unnecessary now.
16:15:35 openstackgerrit Merged openstack/nova master: Add recreate test for unshelve offloaded instance spawn fail https://review.openstack.org/506414
16:18:34 cdent mriedem: rgerganov is doing get_inventory for vmware, already running into some interesting wrinkles
16:19:48 openstackgerrit Eric Fried proposed openstack/nova master: Live Migration sequence diagram https://review.openstack.org/506370
16:20:55 mriedem yufei: you don't need to +1 your own change :)
16:21:05 mriedem yufei: how is that different from the [service_user] group of config options?
16:27:15 bauzas mmm, firefox loves me when I github blame the compute manager file
16:29:35 openstackgerrit Dan Smith proposed openstack/nova master: Add get_node_uuid() helper to ResourceTracker https://review.openstack.org/506730
16:29:36 openstackgerrit Dan Smith proposed openstack/nova master: Move allocation manipulation out of drop_move_claim() https://review.openstack.org/498947
16:29:36 openstackgerrit Dan Smith proposed openstack/nova master: Make allocation cleanup honor new by-migration rules https://review.openstack.org/498948
16:29:37 openstackgerrit Dan Smith proposed openstack/nova master: Pre-create migration object https://review.openstack.org/498950
16:29:37 openstackgerrit Dan Smith proposed openstack/nova master: Revert allocations by migration uuid https://review.openstack.org/498949
16:29:38 openstackgerrit Dan Smith proposed openstack/nova master: Refactor resource tracker to account for migration allocations https://review.openstack.org/506419
16:29:38 openstackgerrit Dan Smith proposed openstack/nova master: WIP: Make migration uuid hold allocations for migrating instances https://review.openstack.org/506420
16:29:39 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Refactor ServerMovingTests for non-move tests https://review.openstack.org/507195
16:29:39 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Add recreate test for unshelve offloaded instance spawn fail https://review.openstack.org/507196
16:29:40 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Remove allocations when unshelve fails on host https://review.openstack.org/507197
16:31:39 yufei @mriedem, thanks for review, keystone session support to re-auth for 401 error no matter what is configred in [service_user], the code is at https://github.com/openstack/python-keystoneclient/blob/b29f478f28c4989156cfe87392cbd308e3f55c1e/keystoneclient/session.py#L432
16:31:59 yufei I’m not sure that whether I understand your question correctly
16:33:53 mriedem yufei: i'm trying to sort out when we can rely on ksa to re-auth and when we need to configure a service user token for re-auth
16:34:10 mriedem for example, if ksa is going to re-auth for us, then why do we even need service user tokens?
16:35:56 yufei the reauth logic was useful when we don’t use keystone session to build ironic client in the old version of nova, I forget the actual version of that, let me find
16:36:32 mriedem yufei: https://specs.openstack.org/openstack/nova-specs/specs/ocata/implemented/use-service-tokens.html if you're not familiar
16:38:01 openstackgerrit Merged openstack/nova master: Remove allocations when unshelve fails on host https://review.openstack.org/506458
16:38:21 openstackgerrit Merged openstack/nova master: Add _wait_for_action_fail_completion to InstanceHelperMixin https://review.openstack.org/506660
16:40:31 yufei I remember that, when user use a token to build keystone client, keystone is unable to reauth for 401 error, but nova service use username and password to build keystone session client to call neutron or neutron, keystone will help to reauth in this way
16:41:54 mriedem maybe edmondsw can reply in the patch, he knows more about ksa than i do
16:42:40 edleafe mriedem: do you have a sucke^H^H^H^H volunteer for API-SIG liaison for Nova?
16:43:06 mriedem edleafe: it's either you or cdent
16:43:18 yufei for live-migration or some other actions, user send a request to nova service with token is context, if token expired, action failed.
16:43:29 edleafe mriedem: it can't be
16:43:34 openstackgerrit Matthew Booth proposed openstack/nova master: libvirt: Don't VIR_MIGRATE_NON_SHARED_INC without migrate_disks https://review.openstack.org/507202
16:43:36 yufei thanks, mriedem
16:43:53 edleafe the liaisons are the ones who review changes *after* the cores have approved them

Earlier   Later