| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2022-06-30 | |||
| 15:12:24 | frickler | the segment test locking the DB leading to retries in other actions | |
| 15:12:59 | sean-k-mooney | oh so you think this trace https://zuul.opendev.org/t/openstack/build/b5c09ce1dbdd42228f5f2928d9df6178/log/controller/logs/screen-q-svc.txt#10722 | |
| 15:13:05 | sean-k-mooney | is caused by the segment test | |
| 15:13:40 | sean-k-mooney | ORM session: SQL execution without transaction in progress, traceback | |
| 15:14:17 | sean-k-mooney | lets check the nova logs and see if there is a rety on our side | |
| 15:14:27 | sean-k-mooney | if not then its an internal neutron issue | |
| 15:14:53 | stephenfin | sean-k-mooney: I'm not sure if that's related to this issue or not | |
| 15:16:50 | stephenfin | actually no, maybe it is. It's an update_port call that's causing the issue | |
| 15:31:49 | opendevreview | Merged openstack/nova stable/xena: reenable greendns in nova. https://review.opendev.org/c/openstack/nova/+/833411 | |
| 16:04:55 | opendevreview | Merged openstack/os-vif master: Check for hybrid plugging in OVS https://review.opendev.org/c/openstack/os-vif/+/839102 | |
| 16:37:44 | opendevreview | Merged openstack/nova master: ignore deleted server groups in validation https://review.opendev.org/c/openstack/nova/+/847001 | |
| 16:54:38 | opendevreview | Merged openstack/nova stable/wallaby: compute: Ensure updates to bdms during pre_live_migration are saved https://review.opendev.org/c/openstack/nova/+/843680 | |
| 16:54:45 | opendevreview | Merged openstack/nova stable/wallaby: fup: Make connection_info returned by CinderFixture unique per attachment https://review.opendev.org/c/openstack/nova/+/844594 | |
| 16:54:51 | opendevreview | Merged openstack/nova stable/wallaby: fup: Assert state of connection_info during LM rollback in func tests https://review.opendev.org/c/openstack/nova/+/844595 | |
| 18:44:31 | opendevreview | Jay Faulkner proposed openstack/nova stable/victoria: [ironic] Minimize window for a resource provider to be lost https://review.opendev.org/c/openstack/nova/+/800873 | |
| #openstack-nova - 2022-07-01 | |||
| 00:33:21 | opendevreview | melanie witt proposed openstack/nova master: libvirt: Report ephemeral encryption traits based on imagebackend https://review.opendev.org/c/openstack/nova/+/826753 | |
| 06:55:42 | kashyap | Has anyone ever tested live migration with vTPM? | |
| 07:47:07 | opendevreview | Balazs Gibizer proposed openstack/nova master: Revert "zuul: Temporarly put Centos9 Stream job non-voting" https://review.opendev.org/c/openstack/nova/+/848352 | |
| 07:47:21 | gibi | bauzas: ^^ I think we can turn it back on | |
| 08:15:09 | bauzas | gibi: I have some concerns with the Centos 9S job I wanted to discuss on the last meeting | |
| 08:15:19 | gibi | we can keep the revert open til next week | |
| 08:15:27 | bauzas | thanks | |
| 08:15:36 | gibi | I just noticed neutron reverted the skip so I thought I propose the revert to not to forget it | |
| 08:15:55 | bauzas | gibi: I'll add it into the agenda | |
| 08:15:59 | gibi | ack | |
| 10:57:41 | manuvakery1 | Hi .. user-data is not getting applied for any instance which is rebuild using a snapshot taken from the same instance | |
| 10:58:06 | manuvakery1 | reported a bug in https://bugs.launchpad.net/nova/+bug/1980511 | |
| 11:04:33 | sean-k-mooney | that looks like a vlaid bug your runing openstack train | |
| 11:05:10 | sean-k-mooney | although i wonder | |
| 11:05:46 | sean-k-mooney | manuvakery1: before you snapshot the instnace did you remove the file that cloud init creates to indicate that it not a first boot | |
| 11:06:22 | sean-k-mooney | if you did not do that then it may not have downloaded an updated copy form teh metadata api | |
| 11:06:43 | sean-k-mooney | manuvakery1: are you using config drive ro jsut the metadata api? | |
| 11:12:27 | manuvakery1 | I am using metadata api | |
| 11:13:35 | manuvakery1 | sean-k-mooney: i have not removed any file which cloud-init created | |
| 11:14:40 | sean-k-mooney | ok then this is likely not a nova bug | |
| 11:15:06 | sean-k-mooney | i dont think cloud init will redonwload the user data file after first boot | |
| 11:16:37 | sean-k-mooney | manuvakery1: https://cloudinit.readthedocs.io/en/latest/topics/cli.html#clean | |
| 11:17:22 | manuvakery1 | but when I tried to rebuild with another snapshot ( created from another server) i see the user-data is getting downloaded | |
| 11:17:58 | sean-k-mooney | the donwloading of the user data is not somethign that nova is invovled with other then servicing it via the metadata api | |
| 11:18:10 | sean-k-mooney | have you checked that the content in the metadata api is correct | |
| 11:18:56 | sean-k-mooney | curl http://169.254.169.254/openstack/latest/user_data | |
| 11:19:23 | sean-k-mooney | if that has the correct content form within the vm then this is not a nova bug | |
| 11:19:52 | manuvakery1 | ok.. I will check that | |
| 11:20:02 | frickler | it may be an per-instance thing, depending on whether the snapshot contains the same instance-id as the rebuilt server | |
| 11:20:31 | sean-k-mooney | ya that too | |
| 11:20:46 | sean-k-mooney | usign the snapshot on a differnt vm might work | |
| 11:20:52 | sean-k-mooney | but not the one it was taken form | |
| 11:21:41 | sean-k-mooney | the instance id is in http://169.254.169.254/2009-04-04/meta-data/instance-id | |
| 11:21:54 | sean-k-mooney | so its somethign the vm can discover and make desiions on | |
| 11:22:34 | manuvakery1 | the content in the metadata api is correct | |
| 11:23:14 | sean-k-mooney | then this is likely because of th instance-id as frickler said or other files created during first boot | |
| 11:27:15 | manuvakery1 | ok.. i will update and close the bug | |
| 11:27:39 | sean-k-mooney | i have it open i can do that now | |
| 11:28:37 | manuvakery1 | ok | |
| 11:28:37 | sean-k-mooney | manuvakery1: there might be a way to configure cloud init to alwasy download the user-data | |
| 11:29:58 | manuvakery1 | i am going to look into that right away .. let me know if any doc/links are with you | |
| 11:31:06 | manuvakery1 | do you think config-drive is going to make any difference? | |
| 11:31:41 | sean-k-mooney | https://cloudinit.readthedocs.io/en/latest/topics/boot.html?highlight=run%20on%20each%20boot#first-boot-determination | |
| 11:32:48 | sean-k-mooney | """By default, cloud-init attempts to determine which case it is running in by checking the instance ID in the cache against the instance ID it determines at runtime. If they do not match, then this is an instance’s first boot; otherwise, it’s a subsequent boot. """ | |
| 11:35:02 | sean-k-mooney | manuvakery1: https://cloudinit.readthedocs.io/en/latest/topics/cli.html#single | |
| 11:35:24 | sean-k-mooney | you might be able to use that to make the user_data moudule run on each boot | |
| 11:36:27 | manuvakery1 | sean-k-mooney: i'll try that . thanks for your response | |
| 11:37:40 | manuvakery1 | Also i had posted another query few days back .. if i take an instance snapshot after resizing to a higher flavor the min_disk property it set to the older flavor disk size .. any hint on this | |
| 11:50:51 | sean-k-mooney | we filter what image porperteis are copied to the new image | |
| 11:51:00 | sean-k-mooney | im not sure if we have already fixed that or not | |
| 11:51:22 | sean-k-mooney | train is getting quite old so there has been a lot of progress over the intervening 5 releases | |
| 11:53:01 | sean-k-mooney | manuvakery1: you could update it via the glance api after the fact as a workaround | |
| 11:58:11 | sean-k-mooney | manuvakery1: ya looking at the dirver code we dont have any code to update it so it would be a new feature | |
| 12:01:21 | sean-k-mooney | manuvakery1: this is the relevent section fo code https://github.com/openstack/nova/blob/b597cb49a8984fbdfd8b9b4d9a692272015d83ed/nova/virt/libvirt/driver.py#L2927-L2993= but updating min_disk is not expected to happen today | |
| 12:03:34 | manuvakery1 | ok | |
| 12:48:44 | Uggla | sean-k-mooney, gibi, bauzas , I looked at instance metadata and especially devices. I attached a volume to my vm and expected to see it in the instance metadata (try to reboot the vm also). However devices field remains empty. Do I miss something ? | |
| 12:49:35 | sean-k-mooney | did you tag the volume | |
| 12:49:39 | bauzas | good question, I dunno | |
| 12:49:54 | sean-k-mooney | we dont generate metadta for the volumes unless you tag them | |
| 12:50:26 | sean-k-mooney | Uggla: https://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/virt-device-role-tagging.html | |
| 12:52:56 | Uggla | sean-k-mooney, ok I'm gonna try. Should we use this tagging mecanisme also with the shares ? | |
| 12:54:36 | Uggla | sean-k-mooney, btw congratulation for your promotion. | |
| 12:55:26 | sean-k-mooney | we shoudl do it automitaclly for the shares i think since we need to provide the share mount thing to the guest | |
| 12:55:38 | sean-k-mooney | the thing that used to mount the filesystem in teh guest | |
| 12:56:38 | sean-k-mooney | Uggla: also thank you i was not aware that was offically announced yet | |
| 13:05:18 | sean-k-mooney | oh today is july 1st so offcially the start of q3 that makes sense | |
| 13:38:54 | Uggla | sean-k-mooney, just to let you know that you were right. Volume is in the instance metadata as soon as tagged. | |
| 13:39:04 | Uggla | sean-k-mooney, thx ! | |
| 13:49:24 | sean-k-mooney | cool | |
| 13:49:38 | sean-k-mooney | we proably could generate some of that metadata by defualt without a tag | |
| 13:49:58 | sean-k-mooney | but we just dont today | |
| 13:50:31 | sean-k-mooney | apprely when you use q35 the pci adresses are also more advisiory rather then somethign you shoudl rely on | |
| 14:02:06 | Uggla | sean-k-mooney, beforing asking I spent time looking at the code, but missed it because it is done in a dictionary comprehension. | |
| 14:02:41 | Uggla | *conditional dictionary comprehension | |
| 14:03:14 | sean-k-mooney | ack you shouldnt worry about asking things like that | |
| 14:03:44 | sean-k-mooney | the channel is for devleopement including expliaing how/why we did things a certen way | |
| 14:04:23 | sean-k-mooney | Uggla: i assume its condtional base in the presence of a tag | |
| 14:04:35 | sean-k-mooney | i have not looked at that in a long time | |
| 14:05:13 | Uggla | sean-k-mooney, I don't worry asking. But I dislike asking things without trying to search on my side first. | |
| 14:05:50 | sean-k-mooney | fair | |
| 14:06:06 | Uggla | sean-k-mooney, fyi tagged_bdms = {_get_device_name(bdm): bdm for bdm in bdms if bdm.tag} in the driver _build_device_metadata method. | |
| 14:06:27 | sean-k-mooney | ah th if bdm.tag | |
| 14:06:31 | sean-k-mooney | ya | |
| 15:24:21 | opendevreview | Merged openstack/nova master: Adds validation for hw machine type in host caps https://review.opendev.org/c/openstack/nova/+/847126 | |
| 18:34:04 | opendevreview | sean mooney proposed openstack/nova master: update default overcommit https://review.opendev.org/c/openstack/nova/+/830829 | |