| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2022-11-18 | |||
| 12:12:12 | opendevreview | sean mooney proposed openstack/nova stable/xena: Add compute restart capability for libvirt func tests https://review.opendev.org/c/openstack/nova/+/864936 | |
| 12:12:14 | opendevreview | sean mooney proposed openstack/nova stable/xena: enable blocked VDPA move operations https://review.opendev.org/c/openstack/nova/+/864937 | |
| 15:04:15 | opendevreview | Rodolfo Alonso proposed openstack/nova master: Bump minimum version of os-vif to 3.1.0 https://review.opendev.org/c/openstack/nova/+/865031 | |
| 15:08:54 | opendevreview | Sylvain Bauza proposed openstack/nova master: Reproducer for bug 1951656 https://review.opendev.org/c/openstack/nova/+/850673 | |
| 15:08:54 | opendevreview | Sylvain Bauza proposed openstack/nova master: Handle mdev devices in libvirt 7.7+ https://review.opendev.org/c/openstack/nova/+/838976 | |
| 15:08:55 | opendevreview | Sylvain Bauza proposed openstack/nova master: Deprecate mdev creation and hardfail on reboot when missing. https://review.opendev.org/c/openstack/nova/+/864418 | |
| 15:09:10 | bauzas | sean-k-mooney: based on your input ^ | |
| 15:09:36 | sean-k-mooney | cool ill take a look shortly | |
| 15:09:50 | sean-k-mooney | by the way backporting to/past xena is a pain | |
| 15:10:01 | sean-k-mooney | due to vmware/sudo-junko | |
| 15:10:17 | sean-k-mooney | and hte fact 2to3 is nolonger a thing | |
| 15:10:32 | bauzas | sean-k-mooney: hold on your comment, I forgot to update the functest due to the comments | |
| 15:10:57 | sean-k-mooney | im im still fighting with running tox so no rush | |
| 15:13:31 | sean-k-mooney | ok got it working finally ... | |
| 15:16:53 | opendevreview | Sylvain Bauza proposed openstack/nova master: Reproducer for bug 1951656 https://review.opendev.org/c/openstack/nova/+/850673 | |
| 15:16:53 | opendevreview | Sylvain Bauza proposed openstack/nova master: Handle mdev devices in libvirt 7.7+ https://review.opendev.org/c/openstack/nova/+/838976 | |
| 15:16:54 | opendevreview | Sylvain Bauza proposed openstack/nova master: Deprecate mdev creation and hardfail on reboot when missing. https://review.opendev.org/c/openstack/nova/+/864418 | |
| 15:17:34 | bauzas | there is goes | |
| 15:28:00 | opendevreview | Anton Kurbatov proposed openstack/nova master: Fix VMs sorting fail in case of comparison two None values https://review.opendev.org/c/openstack/nova/+/865037 | |
| #openstack-nova - 2022-11-19 | |||
| 00:29:51 | opendevreview | Ghanshyam proposed openstack/nova master: DNM: test new defaults https://review.opendev.org/c/openstack/nova/+/864673 | |
| 12:11:46 | opendevreview | Jorhson Deng proposed openstack/nova master: Optimize the small pagesize in numa_fit_instance_to_host https://review.opendev.org/c/openstack/nova/+/864812 | |
| 14:22:07 | opendevreview | Jorhson Deng proposed openstack/nova master: Optimize the small pagesize in numa_fit_instance_to_host https://review.opendev.org/c/openstack/nova/+/864812 | |
| 18:25:29 | frickler | this job last passed 6 months ago, maybe you can drop it to reduce CI load and stop training people to ignore n-v failures? https://zuul.opendev.org/t/openstack/builds?job_name=tempest-integrated-compute-centos-8-stream&project=openstack%2Fnova&skip=0 | |
| 18:28:06 | frickler | I'd also suggest not to run n-v jobs in gate eg. https://zuul.opendev.org/t/openstack/builds?job_name=nova-live-migration-ceph&project=openstack%2Fnova&pipeline=gate&skip=0 | |
| 18:38:02 | opendevreview | Ghanshyam proposed openstack/nova master: DNM: test new defaults https://review.opendev.org/c/openstack/nova/+/864673 | |
| 20:51:21 | opendevreview | Ghanshyam proposed openstack/nova master: DNM: test new defaults https://review.opendev.org/c/openstack/nova/+/864673 | |
| 22:13:32 | opendevreview | Ghanshyam proposed openstack/nova master: Make tenant network policy default to PROJECT_READER_OR_ADMIN https://review.opendev.org/c/openstack/nova/+/865071 | |
| #openstack-nova - 2022-11-20 | |||
| 13:23:29 | opendevreview | Jorhson Deng proposed openstack/nova master: Optimize the small pagesize in numa_fit_instance_to_host https://review.opendev.org/c/openstack/nova/+/864812 | |
| 13:57:01 | opendevreview | Jorhson Deng proposed openstack/nova master: Optimize the small pagesize in numa_fit_instance_to_host https://review.opendev.org/c/openstack/nova/+/864812 | |
| #openstack-nova - 2022-11-21 | |||
| 10:14:54 | opendevreview | Jorhson Deng proposed openstack/nova master: Optimize the small pagesize in numa_fit_instance_to_host https://review.opendev.org/c/openstack/nova/+/864812 | |
| 10:52:58 | opendevreview | Jorhson Deng proposed openstack/nova master: Optimize the small pagesize in numa_fit_instance_to_host https://review.opendev.org/c/openstack/nova/+/864812 | |
| 11:37:59 | opendevreview | Jorhson Deng proposed openstack/nova master: Optimize the small pagesize in numa_fit_instance_to_host https://review.opendev.org/c/openstack/nova/+/864812 | |
| 12:16:58 | admin1 | i have a server with 1 vm .. i want to do maintenance on this server ... when i run openstack server migrate .. it tells me compute host X could not be found .. | |
| 12:17:06 | admin1 | i look into the logs and it has a diff uuid | |
| 12:17:22 | admin1 | if i want to delete this host, it says it has instances, clear it first | |
| 12:17:25 | admin1 | so i am in a bit of catch22 | |
| 12:17:35 | admin1 | cannot fix without migratiing , cannot migrate without fixing | |
| 12:19:20 | sean-k-mooney | admin1: it sounds like you changed the hostname on the server or the host value in the nova.conf | |
| 12:19:47 | admin1 | sean-k-mooney, i have always used the openstack-ansible playbook and never touched a manual setting | |
| 12:19:48 | sean-k-mooney | thats the only way the uuid would change | |
| 12:20:40 | admin1 | how would one fix this ? | |
| 12:29:48 | sean-k-mooney | you need to deterim if the hostname changed first | |
| 12:30:16 | sean-k-mooney | but it likely will need db surgery if it did and you cant set it back | |
| 12:36:40 | admin1 | would changing the resource provider uuid for this hypervisor from old ( non existent) to new one help ? | |
| 12:37:53 | admin1 | i see that the UUID appears in only 1 filed in the compute_nodes tables | |
| 12:46:00 | admin1 | sean-k-mooney, i think the hostname changed from fqdn -> non-fqdn | |
| 12:46:14 | admin1 | hostname remained the same | |
| 13:11:15 | admin1 | sean-k-mooney, how to check own uuid ? | |
| 13:11:19 | admin1 | from the hypervisor | |
| 13:19:43 | opendevreview | Sahid Orentino Ferdjaoui proposed openstack/nova master: compute: enhance compute evacuate instance to support target state https://review.opendev.org/c/openstack/nova/+/858383 | |
| 13:19:44 | opendevreview | Sahid Orentino Ferdjaoui proposed openstack/nova master: api: extend evacuate instance to support target state https://review.opendev.org/c/openstack/nova/+/858384 | |
| 13:21:55 | sahid | o/ gibi sean-k-mooney I have added you change that you were looking for, hope that makes sense | |
| 13:21:58 | sahid | https://review.opendev.org/c/openstack/nova/+/858384/20/nova/api/openstack/compute/evacuate.py#104 | |
| 13:22:09 | sahid | s/you/the | |
| 13:39:07 | sean-k-mooney | sahid: admin1 sorry was on a call downstream. sahid ill try and take a look at yyour change in general later in the week but that section looks like i was expecting so i think that will be fine | |
| 13:39:49 | sean-k-mooney | admin1: that is unforgunete the base way to resolve this issue would be to set teh hostname back to the fqdn | |
| 13:39:49 | admin1 | i have one vm in this i need to migrate .. after that i can just delete /re-initialize it | |
| 13:40:16 | sahid | sean-k-mooney: no worries, thanks a lot for your return | |
| 13:40:16 | sean-k-mooney | can you check the instance.host value for that vm | |
| 13:41:00 | sean-k-mooney | admin1: the instance.host value is ment to match the host value in the nova.conf | |
| 13:41:39 | sean-k-mooney | if you have just one vm the simpleist fix woudl be to set the nova.conf host value on that node to match the instance.host on the vm | |
| 13:41:51 | sean-k-mooney | then you should be able to cold migrate teh vm | |
| 13:42:51 | sean-k-mooney | live migrtate might also work depending on the vm. e.g. if you are using any special feature like sriov or cpu pinning then cold migration has a higher proablity of working | |
| 13:43:14 | admin1 | i was not able to find hostname or name value in nova.conf | |
| 13:43:35 | sean-k-mooney | admin1: if its not set the defautl is socket.gethostname() | |
| 13:44:24 | sean-k-mooney | admin1: https://docs.openstack.org/nova/latest/configuration/config.html#DEFAULT.host | |
| 13:45:31 | sean-k-mooney | admin1: nova does not support changing the hostname because it currpts our db. we have had a bad expirince with customer doing this acidentally of late to the point that we are now working on detecting it and prevent the compute agent form starting when it happens https://review.opendev.org/q/topic:bp%252Fstable-compute-uuids | |
| 13:46:26 | admin1 | Failed to create resource provider record in placement API for UUID 88b9b395-784f-4d78-8497-3d674f7dff64 .. Conflicting resource provider name: h20 already exists .. this is what I have | |
| 13:46:38 | admin1 | so question is where does this UUID come from ? | |
| 13:48:04 | sean-k-mooney | ah yes that makes sesne | |
| 13:48:19 | sean-k-mooney | ok remove the host value | |
| 13:48:28 | sean-k-mooney | and upstea the instnace.host for that one instnace | |
| 13:49:05 | sean-k-mooney | they way the uuid is calulated today is we use the nova.conf host value to look for a compute service record with the same host value | |
| 13:49:37 | sean-k-mooney | *we look for a compute node record with the same host value not comptue service | |
| 13:50:32 | admin1 | so wherever in database, h20 with old UUID appears, i need to just updated it with the new 88b9b395-784f-4d78-8497-3d674f7dff64 uuid ? | |
| 13:55:23 | opendevreview | Alexey Stupnikov proposed openstack/nova stable/wallaby: [stable-only] Use os-brick from source in wallaby https://review.opendev.org/c/openstack/nova/+/865134 | |
| 14:04:08 | sean-k-mooney | admin1: no | |
| 14:04:32 | sean-k-mooney | you should leave teh comptue node alone and update the host value on the one instnace that is affected | |
| 14:04:48 | sean-k-mooney | admin1: presumable its the full fqdn corrently right | |
| 14:04:58 | sean-k-mooney | and the host is not just the hostname not fqdn | |
| 14:05:05 | sean-k-mooney | on the compute node | |
| 14:05:17 | sean-k-mooney | so you need to make them match then migrate it | |
| 14:05:41 | sean-k-mooney | we use the instance.host to determin the rpc endpoint of the compute service that manages it | |
| 14:06:25 | sean-k-mooney | admin1: so if the compute service name change and you have just one vm the shortest way to fix it is update that one instnace and then migrate it | |
| 14:06:30 | admin1 | right .. its full fqdn, but the issue is the current hostname is also not able to register into placement .. it saysFailed to create resource provider record in placement API for UUID 88b9b395-784f-4d78-8497-3d674f7dff64 .. Conflicting resource provider name: h20 already exists | |
| 14:06:54 | admin1 | so just update for this one instance, node to h20 instead of h20.fqdn | |
| 14:07:32 | admin1 | host does show h20 .. node shows h20.fqdn | |
| 14:08:00 | sean-k-mooney | ok so i need you to check a few things all of which shoudl be the same | |
| 14:08:24 | sean-k-mooney | we need to check that the instance.host value and serivce.host value are the same. | |
| 14:08:38 | sean-k-mooney | the hyperviour hostname and placement RP name need to be the same | |
| 14:08:48 | sean-k-mooney | and the compute node uuid and placment uuid need to match | |
| 14:09:09 | sean-k-mooney | and the compute node host value must match the instace.host and service.host values | |
| 14:09:42 | sean-k-mooney | those are the 4 things that need to align. | |
| 14:10:25 | sean-k-mooney | nova does not support changign the hostname or the [DEFAULT]/host value after the agent is first started on a physical server | |
| 14:10:56 | sean-k-mooney | chaning either will currpt both the nova db and create issues in placemnt | |
| 14:11:44 | admin1 | " compute node uuid and placment uuid need to match" - where/how would I see those values ? | |
| 14:11:47 | admin1 | from the db ? | |
| 14:12:09 | sean-k-mooney | yep although you can actuly get them form the api too | |
| 14:12:17 | sean-k-mooney | the placment uuid is jsut in the placement show output | |
| 14:12:31 | sean-k-mooney | the compute node uuid is in the hypervior api if you use a new enough verion | |