Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-10-10
21:16:34 openstackgerrit Chris Dent proposed openstack/nova master: [placement] POST /allocations to set allocations for >1 consumers https://review.openstack.org/500073
21:17:37 edleafe cdent: You mean so they can be in the same microversion? Meh, they're cheap. You can do it in a follow-up patch if you like
21:18:24 coreycb mriedem: it seems to still be initializing the DB from prior nova-manage commands. here's some output of 'list_cells' looping and then running 'discover_hosts' once the cell starts showing up in list_cells output: https://launchpadlibrarian.net/339984742/printscreen
21:18:44 cdent edleafe: no, not worried about the microversion issue, just that somewhere it needs to be marked as we’re not done here. If it is to be done, I’d be inclined to do it in a prior patch.
21:18:53 cdent i’ll leave a comment-to-self
21:19:09 coreycb mriedem: search for 'cell1'
21:20:27 mriedem coreycb: Table 'nova.compute_nodes' doesn't exist
21:20:40 mriedem you can't run discover_hosts or map_instances unti you've synced the dbs on the cells
21:20:55 mriedem meaning, you need to run nova-manage db sync on each cell database
21:21:27 mriedem https://docs.openstack.org/nova/pike/user/cells.html#setup-of-cells-v2
21:22:09 mriedem "At this point, the API database can now find the cell database, and further commands will attempt to look inside. If this is a completely fresh database (such as if you’re adding a cell, or if this is a new deployment), then you will need to run nova-manage db sync on it to initialize the schema."
21:22:16 coreycb mriedem: i think we do that, we basically follow the order of what's in doc/source/cells.rst
21:22:35 coreycb mriedem: it's just a small window and then you'll see at the end of that output discover_hosts works.
21:23:37 mriedem coreycb: so there is a separate process that's calling create_cell and doing the db sync on each cell?
21:28:31 coreycb mriedem: no it's same process, but we don't target the cell on the db sync if that's possible
21:29:15 mriedem nova-manage db sync will sync the current cell db (so cell1 if you're running it on a nova.conf with [database] configured for cell1) and it will query the nova_api database looking for cell0 and sync that as well
21:29:30 mriedem that has to happen before you start discovering hosts in cell1
21:30:53 coreycb mriedem: ok. yes we're doing that. basically we do steps 1-6 of the document you linked, and then if i run list_cells and discover_hosts in a loop, there will be a window (i've not timed it, maybe 30 seconds?) where list_cells returns 'cell1' but discover_hosts fails. then after eventually discover_hosts works.
21:31:27 coreycb mriedem: just curious really if that's normal behavior
21:33:26 openstackgerrit OpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements https://review.openstack.org/511021
21:33:33 mriedem well i can tell from that dump that you're trying to discover_hosts before cell1 db schema is fully sync'ed
21:33:45 mriedem because
21:33:45 mriedem ProgrammingError: (pymysql.err.ProgrammingError) (1146, u"Table 'nova.compute_nodes' doesn't exist")
21:34:21 coreycb mriedem: right. my impression from that is that nova-manage commands aren't necessarily finished when they return. but i'm guessing.
21:34:59 mriedem db sync is synchronoous
21:35:36 openstackgerrit OpenStack Proposal Bot proposed openstack/os-vif master: Updated from global requirements https://review.openstack.org/511035
21:36:38 coreycb mriedem: ok that's good to know. if that's the case it must be something on my end then.
21:36:54 coreycb mriedem: thanks
21:36:59 mriedem np
21:38:41 rm_work hey, when an instance is live-migrated, does nova update the AZ in the database / api response?
21:39:11 rm_work I *assume* yes, but it's one of those things I'd just like to be sure of
21:44:02 mriedem dansmith: remind me, new style migration allocation things, the dest node allocation is held by the instance, and the source node allocation is held by the migration, correct?
21:44:12 dansmith yes
21:44:25 openstackgerrit Jackie Truong proposed openstack/python-novaclient master: Microversion 2.54 - Add trusted_image_certificates https://review.openstack.org/500396
22:08:38 mriedem dansmith: ok comments in https://review.openstack.org/#/c/498949/
22:34:41 openstackgerrit Merged openstack/nova master: Transform aggregate.remove_host notification https://review.openstack.org/396211
23:07:04 openstackgerrit Hongbin Lu proposed openstack/nova master: [WIP] Handle not found error on taking snapshot https://review.openstack.org/511074
23:09:12 openstackgerrit Hongbin Lu proposed openstack/nova master: [WIP] Handle not found error on taking snapshot https://review.openstack.org/511074
23:09:13 openstackgerrit Merged openstack/nova master: Ensure instance can migrate when launched concurrently https://review.openstack.org/506093
#openstack-nova - 2017-10-11
00:27:08 openstackgerrit OpenStack Proposal Bot proposed openstack/os-vif stable/pike: Updated from global requirements https://review.openstack.org/493146
00:39:53 openstackgerrit OpenStack Proposal Bot proposed openstack/os-vif stable/pike: Updated from global requirements https://review.openstack.org/493146
00:52:42 openstackgerrit OpenStack Proposal Bot proposed openstack/os-vif stable/pike: Updated from global requirements https://review.openstack.org/493146
01:12:40 Kevin_Zheng mriedem, hi, for https://review.openstack.org/#/c/509326/
01:16:26 Kevin_Zheng I responded some comment, as for the neutron part, I submitted a RFE and seems Amando is fine with it and they will discuss it at the driver meeting this friday: https://bugs.launchpad.net/neutron/+bug/1718605, how should I put it in the spec?
01:16:27 openstack Launchpad bug 1718605 in neutron "[RFE] Support regex matching when filtering port with IP address" [Wishlist,Triaged] - Assigned to Zhenyu Zheng (zhengzhenyu)
01:17:38 openstackgerrit Jay Pipes proposed openstack/nova master: placement: integrate ProviderTree to report client https://review.openstack.org/415921
01:17:38 openstackgerrit Jay Pipes proposed openstack/nova master: placement: add nested resource providers https://review.openstack.org/377138
01:17:39 openstackgerrit Jay Pipes proposed openstack/nova master: placement: allow filter providers in tree https://review.openstack.org/377215
01:17:39 openstackgerrit Jay Pipes proposed openstack/nova master: placement: adds REST API for nested providers https://review.openstack.org/384807
01:17:40 openstackgerrit Jay Pipes proposed openstack/nova master: placement: update client to set parent provider https://review.openstack.org/385693
02:32:41 openstackgerrit Merged openstack/nova master: Fix minor input items from previous patches https://review.openstack.org/506416
03:39:07 openstackgerrit OpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements https://review.openstack.org/511021
03:41:11 openstackgerrit OpenStack Proposal Bot proposed openstack/os-vif master: Updated from global requirements https://review.openstack.org/511035
05:50:02 openstackgerrit OpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements https://review.openstack.org/511021
05:52:13 openstackgerrit OpenStack Proposal Bot proposed openstack/os-vif master: Updated from global requirements https://review.openstack.org/511035
06:53:48 openstackgerrit Elod Illes proposed openstack/nova master: WIP: Transform scheduler.select_destinations notification https://review.openstack.org/508506
07:22:24 openstackgerrit Elod Illes proposed openstack/nova master: WIP: Transform scheduler.select_destinations notification https://review.openstack.org/508506
07:30:51 openstackgerrit Merged openstack/nova master: Target context for build notification in conductor https://review.openstack.org/509967
08:28:13 openstackgerrit Surya Seetharaman proposed openstack/nova master: nova-manage map_instances is not using the cells info from the API database https://review.openstack.org/510844
08:33:41 openstackgerrit Huan Xie proposed openstack/nova master: [WIP] VGPU: Define vgpu related resources https://review.openstack.org/511180
08:36:08 openstackgerrit Huan Xie proposed openstack/nova master: [WIP] VGPU: Define vgpu related resources https://review.openstack.org/511180
08:41:39 openstackgerrit Huan Xie proposed openstack/nova master: [WIP] VGPU: Define vgpu related resources https://review.openstack.org/511180
08:41:50 openstackgerrit sahid proposed openstack/nova master: libvirt: set emulator threads on isolated host cpus https://review.openstack.org/510897
08:45:08 openstackgerrit Huan Xie proposed openstack/nova master: [WIP] VGPU: Define vgpu related resources https://review.openstack.org/511180
08:49:54 kashyap mdbooth: Hi there, when you get a moment:
08:50:06 mdbooth kashyap: Wassup?
08:50:30 kashyap mdbooth: Do you recall top off your head, if Nova sets any QEMU migration parameters? Or do we just rely on libvirt? Let me give an example of what I mean
08:51:28 kashyap Take a look at the ouput of this command: virsh qemu-monitor-command f26-devstack --pretty '{"execute":"query-migrate-parameters"}'
08:51:32 kashyap http://paste.openstack.org/show/623292/
08:52:29 stephenfin kashyap: to the best of my knowledge, if it can't be done via libvirt then we don't do it
08:52:56 mdbooth kashyap: nova never talks to qemu directly if that's what you mean?
08:53:00 kashyap stephenfin: Yep. I know we don't "do direct QEMU", that's the whole frelling point of explicitly modelling in libvirt XML aspects
08:53:14 mdbooth In fact, I don't think qemu allows more than one monitor, does it?
08:53:18 mdbooth i.e. it's not even possible
08:54:14 kashyap mdbooth: I know Nova never talks to QEMU
08:54:22 kashyap mdbooth: And QEMU does allow more than one monitor.
08:54:45 mdbooth kashyap: I didn't understand the question, then
08:54:57 kashyap mdbooth: Let me give a bit more context:
08:55:06 mdbooth Nova does set some of those values explicitly, via libvirt
08:55:28 kashyap mdbooth: What made me ask the question is, one of the QEMU migration devs asked me this morning what migration parameters Nova uses
08:56:27 kashyap mdbooth: You (libvirt could) set migration parameters via `migrate-set-parameters` QMP command, _except_ this "xbzrle-cache-size"
08:56:49 kashyap And they were wondering does Nova use it in anyway. So I went digging...
08:57:15 mdbooth Yes, absolutely. Nova tweaks a bunch of those.
09:00:38 kashyap mdbooth: Meanwhile, here's an example for multiple monitors:
09:00:42 kashyap $ qemu-system-x86_64 [...] -monitor stdio -qmp unix:./qmp-sock,server,nowait
09:01:00 kashyap HMP (Human Monitor Protocol) on 'stdio', and QMP on a UNIX socket.
09:06:19 openstackgerrit Andrey Volkov proposed openstack/osc-placement master: [WIP] CLI for resource classes https://review.openstack.org/511182
09:06:19 openstackgerrit Andrey Volkov proposed openstack/osc-placement master: [WIP] RP list: member_of and resources parameters https://review.openstack.org/511183
09:06:49 openstackgerrit Alex Xu proposed openstack/nova master: [DNM] Add _get_provider_ids_with_any_trait https://review.openstack.org/511184
09:06:49 openstackgerrit Alex Xu proposed openstack/nova master: [DNM] placement: add traits support in the AllocationCandidates object https://review.openstack.org/511185
09:12:14 openstackgerrit sahid proposed openstack/nova-specs master: update: introducing isolate emulthreads on host https://review.openstack.org/511188
09:42:53 openstackgerrit jichenjc proposed openstack/nova master: Remove duplicate error info https://review.openstack.org/510719
09:44:03 openstackgerrit Takashi NATSUME proposed openstack/nova master: Enable cold migration with target host(1/2) https://review.openstack.org/408955
09:48:00 openstackgerrit Takashi NATSUME proposed openstack/nova master: Enable cold migration with target host(1/2) https://review.openstack.org/408955
09:48:59 openstackgerrit Takashi NATSUME proposed openstack/nova master: Enable cold migration with target host(2/2) https://review.openstack.org/408964
09:59:20 openstackgerrit Hironori Shiina proposed openstack/nova master: ironic: Support resize and cold migration https://review.openstack.org/500677
09:59:58 openstackgerrit jichenjc proposed openstack/nova master: Remove duplicate error info https://review.openstack.org/510719
10:00:50 openstackgerrit Takashi NATSUME proposed openstack/nova master: List/show all server migration types (1/2) https://review.openstack.org/430608
10:06:42 lyarwood mdbooth: so the stable rescue spec LGTM. I've been umming and ahhing about the sync vs async failure paths for a while, the only alternative I can see is making the rescue_instance rpc cast a call but that's a horrid change.
10:07:13 mdbooth lyarwood: On a call. Sync up in a bit.

Earlier   Later