Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-07-19
21:03:47 artom efried, \o/
21:04:23 efried dansmith: we got this one all ready for you https://review.opendev.org/#/c/671471/
21:23:05 openstackgerrit Eric Fried proposed openstack/nova master: Remove @safe_connect from _delete_provider https://review.opendev.org/671866
21:41:11 openstackgerrit Eric Fried proposed openstack/nova master: Remove @safe_connect from _delete_provider https://review.opendev.org/671866
22:03:40 openstackgerrit Eric Fried proposed openstack/nova master: Disambiguate logs in delete_allocation_for_instance https://review.opendev.org/671869
22:03:54 efried o/
22:46:11 openstackgerrit Dustin Cowles proposed openstack/nova master: Use SDK for node.list https://review.opendev.org/656027
22:46:12 openstackgerrit Dustin Cowles proposed openstack/nova master: Use SDK for validating instance and node https://review.opendev.org/656028
22:46:12 openstackgerrit Dustin Cowles proposed openstack/nova master: Use SDK for setting instance id https://review.opendev.org/659690
22:46:13 openstackgerrit Dustin Cowles proposed openstack/nova master: Use SDK for add/remove instance info from node https://review.opendev.org/659691
22:46:13 openstackgerrit Dustin Cowles proposed openstack/nova master: Use SDK for getting network metadata from node https://review.opendev.org/670213
#openstack-nova - 2019-07-20
02:17:34 openstackgerrit Merged openstack/nova master: Optimize SchedulerReportClient.delete_resource_provider https://review.opendev.org/657070
03:26:43 openstackgerrit Merged openstack/nova master: Exit 1 when db sync runs before api_db sync https://review.opendev.org/665626
07:21:08 openstackgerrit zhangyangyang proposed openstack/python-novaclient master: Bump the openstackdocstheme extension to 1.20 https://review.opendev.org/671887
17:42:14 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Run 'tempest-ipv6-only' job in gate https://review.opendev.org/671900
#openstack-nova - 2019-07-21
01:40:40 openstackgerrit Merged openstack/python-novaclient master: Bump the openstackdocstheme extension to 1.20 https://review.opendev.org/671887
13:52:24 openstackgerrit Stephen Finucane proposed openstack/nova master: objects: Rename 'nova.objects.instance_numa_topology' https://review.opendev.org/671789
13:52:25 openstackgerrit Stephen Finucane proposed openstack/nova master: libvirt: Remove unnecessary try-catch around 'getCPUMap' https://review.opendev.org/671790
13:52:25 openstackgerrit Stephen Finucane proposed openstack/nova master: objects: Remove legacy '_from_dict' functions https://review.opendev.org/537414
13:52:26 openstackgerrit Stephen Finucane proposed openstack/nova master: claims: Remove useless caching https://review.opendev.org/671791
13:52:26 openstackgerrit Stephen Finucane proposed openstack/nova master: Add '[compute] cpu_dedicated_set' option https://review.opendev.org/671792
13:52:27 openstackgerrit Stephen Finucane proposed openstack/nova master: libvirt: Start reporting PCPU inventory to placement https://review.opendev.org/671793
13:52:27 openstackgerrit Stephen Finucane proposed openstack/nova master: libvirt: Start reporting 'pcpus' to the resource tracker https://review.opendev.org/671794
13:52:28 openstackgerrit Stephen Finucane proposed openstack/nova master: trivial: Rename exception argument https://review.opendev.org/671795
13:52:28 openstackgerrit Stephen Finucane proposed openstack/nova master: trivial: Remove unused function parameter https://review.opendev.org/671796
13:52:29 openstackgerrit Stephen Finucane proposed openstack/nova master: Remove 'hardware.get_host_numa_usage_from_instance' https://review.opendev.org/671797
13:52:29 openstackgerrit Stephen Finucane proposed openstack/nova master: Remove 'hardware.host_topology_and_format_from_host' https://review.opendev.org/671798
13:52:30 openstackgerrit Stephen Finucane proposed openstack/nova master: Remove 'hardware.instance_topology_from_instance' https://review.opendev.org/671799
13:52:30 openstackgerrit Stephen Finucane proposed openstack/nova master: WIP: hardware: Differentiate between shared and dedicated CPUs https://review.opendev.org/671800
13:52:31 openstackgerrit Stephen Finucane proposed openstack/nova master: Add support translating CPU policy extra specs, image meta https://review.opendev.org/671801
#openstack-nova - 2019-07-22
01:31:50 brinzhang lyarwood: After you review this spec[1] 5/20, now this spec wait for your update, when do you have time to do it? [1]https://review.opendev.org/#/c/580336/
05:47:52 alex_xu efried: bauzas Luyao found a problem https://review.opendev.org/#/c/671222/4/nova/virt/libvirt/device.py@119, I guess we should use the migration as the consumer when migrate, just like we did for the placement
07:32:57 openstackgerrit wangwei1 proposed openstack/nova master: fix spelling error in nova/api/validation/__init__.py https://review.opendev.org/669244
08:03:05 bauzas alex_xu: ack, I'll need to look
09:00:57 alex_xu bauzas: thanks, after take a look the detail, I don't think the migration uuid works. in the end, I'm thinking passing the flavor id to the claim/unclaim_for_instance. Then claim/unclaim based on the (instance_uuid, flavor_id)
09:26:00 openstackgerrit **** proposed openstack/nova master: Nova: node should be deleted when last service is deleted https://review.opendev.org/671731
11:29:23 sean-k-mooney stephenfin: when you get back form lunch we have killed the cacheing schduler so maybe we should delete the core, ram and disk filters then nuke the filed for the resouce tracker?
11:30:20 sean-k-mooney it would be a change to the hyperviors api but you could always proxy the data from placement.
11:37:25 openstackgerrit **** proposed openstack/nova master: Nova: node should be deleted when last service is deleted https://review.opendev.org/671731
11:41:04 sean-k-mooney johnthetubaguy: lyarwood can ye take a look at this stable backport https://review.opendev.org/#/c/671532/ if ye have time
11:57:38 openstackgerrit Alex Xu proposed openstack/nova master: Add the virt driver interface for claim and unclaim the devices https://review.opendev.org/670782
11:57:39 openstackgerrit Alex Xu proposed openstack/nova master: Moves the allocation retrieving early https://review.opendev.org/670783
11:57:39 openstackgerrit Alex Xu proposed openstack/nova master: Calling the virt driver's claim/unclaim_for_instance in resource tracker https://review.opendev.org/670784
11:57:40 openstackgerrit Alex Xu proposed openstack/nova master: Add DeviceManager to the libvirt virt driver https://review.opendev.org/671388
11:57:40 openstackgerrit Alex Xu proposed openstack/nova master: Populates the existing mediated devices in the libvirt device manager https://review.opendev.org/670787
11:57:41 openstackgerrit Alex Xu proposed openstack/nova master: Using the claim/unclaim_for_instance for mdevs https://review.opendev.org/671222
11:57:41 openstackgerrit Alex Xu proposed openstack/nova master: Adds functional test for creating the instance with vgpus https://review.opendev.org/671398
12:01:44 ygk_12345 hi all
12:02:53 ygk_12345 some of the vms on the compute node are eating too much of memory. Can someone point me to any links for finding out what is happening in those vms from the hypervisro prespective ?
12:18:01 TheJulia sean-k-mooney: I never get enough sleep and if I don't take meds It would be even worse... so I always needssleep :)
12:18:33 mdbooth ygk_12345: You're unlikely to find the right expertise here for that. Assuming you're using libvirt/kvm you'll want to talk to a libvirt/kvm forum. Or your vendor...
12:19:08 ygk_12345 mdbooth is there any irc channel for kvm in general ?
12:23:15 openstackgerrit sean mooney proposed openstack/nova master: Libvirt: add support for vPMU configuration. https://review.opendev.org/671338
12:23:27 mdbooth ygk_12345: Try #virt on irc.oftc.net
12:23:40 ygk_12345 mdbooth ok thanks
12:24:32 sean-k-mooney ygk_12345: how much memory are they consuming over what you execpt
12:25:36 ygk_12345 sean-k-mooney what I observer is in general the cpu usage against those qemu-kvm processes is shooting above 150 sometimes and so I am suspecting them. If I shut them down it is normal again
12:26:20 sean-k-mooney ygk_12345: are you seeing high cpu usage, or high memory usage or both
12:26:26 ygk_12345 both
12:26:30 ygk_12345 sometimes
12:26:49 ygk_12345 how to find out whats exactly happening at those times ?
12:28:30 sean-k-mooney you would likely need to look at the guest journal/logs to be honest but you could check the host for kernel errors
12:28:41 ygk_12345 hmm
12:28:48 sean-k-mooney e.g. check if the kernel is soft locking a core
12:29:12 sean-k-mooney that could be caused by the guest soft locking up
12:29:21 ygk_12345 ok
12:29:45 sean-k-mooney are you currently using q35 machine type or are your still using the pc i440 machine type
12:30:04 ygk_12345 we use dell with latest intel
12:30:07 sean-k-mooney we have seen that q35 while more modern tends to use more memory
12:30:17 ygk_12345 oh ok
12:49:24 boxiang_ https://review.opendev.org/#/c/649963/ and https://review.opendev.org/#/c/651969/ Can someone help review these two patches? thanks :)
13:00:06 efried alex_xu: I don't understand. Why would we not use the migration consumer to migrate?
13:08:57 sean-k-mooney gibi_off: not sure if your are on vaction/PTO this week but just an fyi that i should have adressed your feedback in https://review.opendev.org/671338
13:14:11 openstackgerrit Matt Riedemann proposed openstack/nova stable/stein: Restore RT.old_resources if ComputeNode.save() fails https://review.opendev.org/672038
13:14:59 sean-k-mooney efried: i am also confused. isnt a migration consumer proposed for use when doing cold and live migrations to hold the allocation on the second node
13:15:23 efried sean-k-mooney: He's off for two weeks. It looks like the changes are minor; if so another core will probably proxy his +2.
13:15:30 efried sean-k-mooney: that's my confusion as well.
13:15:42 efried perhaps we only use the migration consumer for live migration?
13:15:44 efried I don't know.
13:15:52 sean-k-mooney efried: ah ok good to know.
13:15:54 mriedem they're used for both cold and live
13:16:13 mriedem just not evacuate
13:16:19 sean-k-mooney mriedem: o/ welcome back, have a good vaction?
13:16:25 mriedem thanks, yeah
13:16:47 sean-k-mooney we dont use them for evac because we dont use them in rebuild
13:17:17 mriedem um
13:17:23 mriedem that's not really the reason
13:17:48 mriedem it's because for evac the source compute service is down, so it's not really the same flow for how the migration-based allocations are handled at the end
13:18:06 mriedem we have migration records for evac
13:18:25 alex_xu efried: if we use migration as consumer for the dst, when we confirm the resize, there isn't a call to dest host to change migration consumer back to instance consumer
13:18:35 sean-k-mooney right however event with a down host we could still free the allocation on the source node in placment
13:18:52 mriedem sean-k-mooney: we do that when the source compute service comes back up
13:19:12 efried alex_xu: oh, you're talking about swapping which side gets which consumer?
13:19:12 alex_xu efried: if we use migration as consumer for the src, we will call the the dest host to claim the device, so we have no way to change the src host's claim consumer to migration.
13:19:14 sean-k-mooney yes that is true.
13:19:18 alex_xu efried: yes
13:19:34 mriedem we *could* use migration-based consumers for evac, i just don't think it was really a priority when dansmith implemented that blueprint
13:20:04 mriedem i'm obviously walking into something late here though
13:20:09 mriedem who is out for 2 weeks?
13:20:13 efried gibi_off:

Earlier   Later