| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2020-05-26 | |||
| 18:41:50 | sean-k-mooney | gibi: by the way i belive the issue in https://bugs.launchpad.net/neutron/+bug/1880389 is i think neutron is telling us the network backend is wired up via the netwrok-vif-plugged event before it acctully add the subport to the linux bridge that connects it to the physical network | |
| 18:42:05 | sean-k-mooney | in that specific case before it add the vlan port to the bridge | |
| 18:42:18 | sean-k-mooney | although the same would happen for vxlan | |
| 18:42:37 | sean-k-mooney | i added neutorn too the bug | |
| 18:43:13 | sean-k-mooney | i treied repoducing it today and on my lightly loaded system i did not see it | |
| 18:43:40 | sean-k-mooney | but i think im just winning the race and the l2 agent happens to wire it up quickly | |
| 18:43:57 | sean-k-mooney | i might look into this a littel more tomorow to confirm but im calling it a day o/ | |
| 19:57:32 | openstackgerrit | Hervé Beraud proposed openstack/nova master: Cap jsonschema 3.2.0 as the minimal version https://review.opendev.org/730960 | |
| 19:57:32 | openstackgerrit | Hervé Beraud proposed openstack/nova master: Cap jsonschema 3.2.0 as the minimal version https://review.opendev.org/730960 | |
| 21:33:38 | melwitt | mnaser: hey, I know and I updated it to call out the conflicting change-ids in the commit message. so probably someone else like dansmith should review | |
| 21:38:08 | dansmith | melwitt: mnaser: done | |
| 21:40:26 | melwitt | thanks | |
| 22:09:29 | mnaser | melwitt: oh cool, sorry, didnt notice that. thanks dansmith :) | |
| #openstack-nova - 2020-05-27 | |||
| 00:15:55 | openstackgerrit | Merged openstack/nova stable/stein: Make quotas respect instance_list_per_project_cells https://review.opendev.org/729088 | |
| 01:08:37 | ikla | every time I try to boot a gpu flavor, the instance gets stuck in spawning and it looks like it loses connectivity to nova-conductor. Any ideas? | |
| 03:05:41 | ikla | I have a dual port card that I created 4 VF's per port for. How do I tell which physical port it's coming from when I do pci passthrough? I have an alias for it but it covers the entire card since I only use vendor_id and product_id. | |
| 07:53:05 | openstack | Launchpad bug 1880389 in OpenStack Compute (nova) "lost net connection when live migration" [Undecided,New] | |
| 07:53:05 | gibi | sean-k-mooney: thanks for looking at https://bugs.launchpad.net/nova/+bug/1880389 I hope neutron folks will confirm what you found and then we can close the nova side of that bug | |
| 07:59:21 | bauzas | gibi: fwiw, I'm on PTO today | |
| 08:00:00 | gibi | bauzas: ack, have a nice PTO | |
| 08:00:54 | gibi | my day is also a bit hectic | |
| 08:25:05 | brinzhang | gibi, dansmith: good morning^ | |
| 08:34:43 | luyao | stephenfin: could you have a look at https://review.opendev.org/#/c/726284/, it's a tiny spec update. | |
| 08:38:00 | brinzhang | gibi, dansmith: do you have time to review sean-k-mooney's rebuild/evacuate cyborg manager accelerators instance patch? https://review.opendev.org/#/c/715326/ | |
| 08:38:06 | brinzhang | we updated it and submit the patch to support shelve/unshelve, suspend/resume instance, and that rebased on this patch. | |
| 08:58:21 | lyarwood | https://review.opendev.org/#/q/topic:bug/1550919+status:open - I'm looking for reviews on this bugfix series if anyone has time today, I'll raise it tomorrow in the stuck reviews section if I'm around, stephenfin can you raise it if I'm AFK? | |
| 09:09:54 | lyarwood | `It's not for reviews that just haven't had attention, except for exceptional cases.` - actually ignore my last, stuck reviews isn't for that. | |
| 09:19:00 | lyarwood | elod: https://review.opendev.org/#/c/730310/ - would you mind hitting this today if you have a chance? | |
| 09:20:59 | elod | lyarwood: looking | |
| 09:21:33 | lyarwood | many thanks | |
| 09:40:50 | openstackgerrit | Stephen Finucane proposed openstack/nova master: trivial: Remove log translations https://review.opendev.org/728966 | |
| 09:41:31 | openstackgerrit | Stephen Finucane proposed openstack/nova master: hacking: Modify checks for translated logs https://review.opendev.org/728967 | |
| 09:41:37 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove hooks https://review.opendev.org/728968 | |
| 09:41:43 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Remove workaround for unsupported libvirt version https://review.opendev.org/729656 | |
| 09:45:18 | openstackgerrit | Liang Fang proposed openstack/nova master: Add volume local cache support https://review.opendev.org/663542 | |
| 10:17:35 | openstackgerrit | vinay harsha mitta proposed openstack/nova stable/rocky: Documentation Bug with PLacement API port https://review.opendev.org/731141 | |
| 11:01:13 | openstack | bug 1869050 in OpenStack Compute (nova) rocky "migration of anti-affinity server fails due to stale scheduler instance info" [Low,In progress] https://launchpad.net/bugs/1869050 - Assigned to Balazs Gibizer (balazs-gibizer) | |
| 11:01:13 | openstackgerrit | Merged openstack/nova stable/rocky: Reproduce bug 1869050 https://review.opendev.org/730343 | |
| 11:42:52 | brinzhang | cyborg | |
| 12:13:17 | openstackgerrit | Merged openstack/nova master: Remove eventlet hub workaround for monotonic clock https://review.opendev.org/644237 | |
| 12:47:00 | openstackgerrit | Merged openstack/nova stable/train: fix scsi disk unit number of the attaching volume when cdrom bus is scsi https://review.opendev.org/730310 | |
| 12:47:13 | openstack | bug 1550919 in OpenStack Compute (nova) "[Libvirt]Evacuate fail may cause disk image be deleted" [Medium,In progress] https://launchpad.net/bugs/1550919 - Assigned to Lee Yarwood (lyarwood) | |
| 12:47:13 | openstackgerrit | Merged openstack/nova master: Add functional test for bug 1550919 https://review.opendev.org/631294 | |
| 12:47:17 | openstackgerrit | Merged openstack/nova master: libvirt: Don't delete disks on shared storage during evacuate https://review.opendev.org/578846 | |
| 12:56:43 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add emulated TPM support to Nova https://review.opendev.org/631363 | |
| 12:56:44 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add support for resize and cold migration of emulated TPM files https://review.opendev.org/639934 | |
| 13:04:50 | jkulik | does anybody know of a solution for giving out quota per AZ instead of per region? from what I read here https://wiki.openstack.org/wiki/Quotas quotas are supposed to help against resource exhaustion, but if all projects only use one AZ, that AZ will still be full while there are lots of resources left in the region | |
| 14:01:25 | sean-k-mooney | jkulik: that is not really how quotas are ment to work | |
| 14:04:57 | jkulik | sean-k-mooney, where can I read up on how quotas are meant to work? | |
| 14:27:21 | sean-k-mooney | jkulik: the general quota docs are here https://docs.openstack.org/nova/latest/user/quotas.html, https://specs.openstack.org/openstack/nova-specs/specs/train/implemented/count-quota-usage-from-placement.html descirbs some of the ground work that will be use with unified limits https://specs.openstack.org/openstack/nova-specs/specs/ussuri/approved/unified-limits-nova.html which will fully replace nova | |
| 14:27:24 | sean-k-mooney | current quota system | |
| 14:27:48 | sean-k-mooney | jkulik: unified limits is not implemeted yet but it should hopefully get completed in victoria | |
| 14:28:30 | sean-k-mooney | jkulik: but the tl;dr is that qutoa and limits are not ment to be specific to host,aggreate aviablity zones | |
| 14:29:10 | sean-k-mooney | the are ment to extend across the entire cloud and are limits on the amount of resources that a tenant can consome form a cloud | |
| 14:29:56 | sean-k-mooney | placement of resouces in speific regions or hosts is a sperate concern of the secudler and placment services | |
| 14:30:10 | sean-k-mooney | both are entirely seperate concerns in openstack | |
| 14:32:55 | sean-k-mooney | johnthetubaguy: by the way are you planning to reporpose https://review.opendev.org/#/c/602201/ | |
| 14:34:09 | sean-k-mooney | johnthetubaguy: i think that is more or less a given that we will continue this for victoria but just wondering if your going to submit the review. to add it back for victoria | |
| 14:53:02 | jkulik | sean-k-mooney, thanks for clarifying that. we're misusing quota then, as we're doing invoices based on quota and then people end up being unable to use what they pay for, if the AZ gets full. | |
| 14:55:39 | sean-k-mooney | jkulik: is see quotas are seperate form the idea of reservations | |
| 14:56:06 | sean-k-mooney | e.g. a quota is a limit on how much reouces you can use not a reservation of resouces for you to use | |
| 14:59:21 | jkulik | sean-k-mooney, for billing purposes, both would have to be the same. the user should not be able to user more than they payed for. but I get you point, I think. | |
| 15:01:47 | sean-k-mooney | jkulik: but they would not normally be usign all there quota | |
| 15:02:04 | sean-k-mooney | so its only the same if they are billed by available capasity instead of usage | |
| 15:02:32 | sean-k-mooney | most clouds bill by usage not capasity unless its a dedicated install | |
| 15:02:40 | jkulik | sean-k-mooney, yeah, they're billed what they could in theory consume. not the actual consumption. | |
| 15:07:09 | sean-k-mooney | jkulik: ya so in that modely you would tipleicaly use one of the feature that allows you to reserve hosts for a given tenat and dedicate those hosts to that tenant of a group of tenatns | |
| 15:08:55 | ikla | How do I tell which virtual interface passes through to nova compute instance. I have a alias for a dual-port card that has 4 VF's per port totalling 8. How do I know which port is passed through since they are on different networks. My alias name is eth10g and it can do say eth10g:8 which would be all 8 VF's. | |
| 15:09:52 | sean-k-mooney | ikla: well as a starting point pci pasthough is not intended for use with nics that will connect to neutron netwroks | |
| 15:10:00 | sean-k-mooney | ikla: specificly via alias | |
| 15:10:22 | sean-k-mooney | but you can find out wich was allocated in two ways | |
| 15:10:56 | sean-k-mooney | either you can check the pci_devices tables | |
| 15:11:08 | sean-k-mooney | in the cell db | |
| 15:11:23 | sean-k-mooney | or you can look at the instance domain xml | |
| 15:11:27 | sean-k-mooney | so do virsh list | |
| 15:16:16 | openstackgerrit | Merged openstack/os-traits master: CPU: add traits for new AVX512 support https://review.opendev.org/728576 | |
| 15:36:16 | ikla | sean-k-mooney, they wont be used with neutron | |
| 15:37:34 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/ussuri: libvirt: Don't delete disks on shared storage during evacuate https://review.opendev.org/731248 | |
| 15:37:34 | openstack | bug 1550919 in OpenStack Compute (nova) "[Libvirt]Evacuate fail may cause disk image be deleted" [Medium,Fix released] https://launchpad.net/bugs/1550919 - Assigned to Lee Yarwood (lyarwood) | |
| 15:37:34 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/ussuri: Add functional test for bug 1550919 https://review.opendev.org/731247 | |
| 15:43:01 | ikla | Please ensure all devices within the iommu_group are bound to their vfio bus driver | |
| 15:43:06 | ikla | anyone seen that before? | |
| 15:43:23 | sean-k-mooney | yes | |
| 15:43:45 | sean-k-mooney | so you cant have diffenret VF bound to different direvers if they are in the same iommu group | |
| 16:37:26 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/train: libvirt: Don't delete disks on shared storage during evacuate https://review.opendev.org/731273 | |
| 16:37:26 | openstack | bug 1550919 in OpenStack Compute (nova) "[Libvirt]Evacuate fail may cause disk image be deleted" [Medium,Fix released] https://launchpad.net/bugs/1550919 - Assigned to Lee Yarwood (lyarwood) | |
| 16:37:26 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/train: Add functional test for bug 1550919 https://review.opendev.org/731272 | |
| 17:13:31 | openstackgerrit | Merged openstack/nova-specs master: Add rbd-glance-multistore spec https://review.opendev.org/730126 | |
| 17:20:09 | stephenfin | lyarwood: before you finish up, could you hit https://review.opendev.org/728966 again? | |
| 17:25:32 | lyarwood | stephenfin: ack done | |
| 17:26:26 | stephenfin | thanks | |
| 18:15:56 | openstackgerrit | Lin Yang proposed openstack/nova master: Add traits for new AVX512 CPU feature https://review.opendev.org/731288 | |
| 18:58:08 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/stein: fix scsi disk unit number of the attaching volume when cdrom bus is scsi https://review.opendev.org/730311 | |
| 18:58:22 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/rocky: fix scsi disk unit number of the attaching volume when cdrom bus is scsi https://review.opendev.org/730312 | |
| 18:58:35 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/queens: fix scsi disk unit number of the attaching volume when cdrom bus is scsi https://review.opendev.org/730313 | |
| 19:34:05 | openstackgerrit | Merged openstack/nova master: trivial: Remove log translations https://review.opendev.org/728966 | |
| 19:46:22 | openstackgerrit | Merged openstack/nova-specs master: Fix hacking min version to 3.0.1 https://review.opendev.org/730162 | |
| 19:46:23 | openstackgerrit | Merged openstack/nova-specs master: [Community goal] Update contributor documentation https://review.opendev.org/722393 | |
| 21:01:26 | openstackgerrit | Lee Yarwood proposed openstack/nova master: Include removal of ephemeral backing files in the image cache manager https://review.opendev.org/689422 | |
| 21:04:24 | openstackgerrit | Merged openstack/nova master: Remove compatibility check from migrate_data https://review.opendev.org/729949 | |