| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2023-10-13 | |||
| 13:55:11 | bauzas | I don't know what happened to me when I started writing that code but I regret every single day that nova supports creating mdevs on the fly | |
| 13:55:48 | sean-k-mooney | the creation isnt relaly the problem. the fact we dont store them in the db is where a lot of the pain comse form | |
| 13:55:58 | sean-k-mooney | since we cant recreate them as needed with the same values | |
| 13:56:48 | sean-k-mooney | bauzas: so i guess for now to get soemthign workign we can precreate them in the ci job | |
| 13:57:16 | bauzas | I'm just out of band for now | |
| 13:57:17 | sean-k-mooney | but i think we can just make this work even without that with minimal extra complexity | |
| 13:57:25 | bauzas | the customer issue went back | |
| 13:57:49 | bauzas | and I also have a lot of back-and-forths internally with people afraid of the situation | |
| 13:58:13 | bauzas | so I basically punted to continue working on mtty for today | |
| 13:58:29 | bauzas | but last time I tried on my vm, mdev creation was working | |
| 13:58:43 | bauzas | what I haven't tried yet was to download my patch and test it | |
| 13:58:50 | bauzas | with precreated mdevs | |
| 13:59:45 | dansmith | sean-k-mooney: ++ | |
| 14:00:29 | dansmith | honestly I think creating on the fly makes plenty of sense, | |
| 14:00:43 | dansmith | we just need a persistence mechanism, which is either libvirt or us | |
| 14:01:09 | sean-k-mooney | we can perssitee them in the resouce table without any db modification | |
| 14:01:13 | sean-k-mooney | if we want too | |
| 14:01:17 | sean-k-mooney | or have libvirt do it | |
| 14:01:25 | dansmith | I don't know that not being able to have libvirt persist mtty is really a problem.. like if that works for real devices and the deal is you just can restart a dev box with mtty fake devices, that's not too bad | |
| 14:01:26 | sean-k-mooney | both work | |
| 14:01:53 | bauzas | the libvirt feature already exists | |
| 14:01:55 | bauzas | https://libvirt.org/drvnodedev.html#example-of-a-mediated-device | |
| 14:01:59 | sean-k-mooney | i know | |
| 14:02:08 | bauzas | let's just use it | |
| 14:02:11 | sean-k-mooney | which is why we want to use it. | |
| 14:02:23 | sean-k-mooney | but im just saying this is a very solvable problem | |
| 14:02:43 | bauzas | I'm not particulatly in favor of adding another datastructure in nova for something we can let the hypervisor handle | |
| 14:03:03 | bauzas | and yet again a sync problem between the hypervisor and us | |
| 14:03:19 | bauzas | (but that's MHO) | |
| 14:03:38 | bauzas | again, priorities matter and that's on my scope | |
| 14:03:53 | bauzas | I just have a limited 70hours bandwidth this week | |
| 14:04:00 | dansmith | yep, agree we should not build data structures if we don't have to | |
| 14:04:33 | sean-k-mooney | the reason we diding in the frist place was just because this was all ment to be tracked via placement | |
| 14:05:24 | sean-k-mooney | but we didnt actully impelmetnet it in such a way that would work since we have no encodign of which mdev was for which allocation | |
| 14:05:39 | sean-k-mooney | but sure sound like we have a short term adn medium path forward | |
| 14:05:51 | sean-k-mooney | shortterm just preacllocate the mdevs | |
| 14:05:54 | sean-k-mooney | in the job | |
| 14:06:15 | sean-k-mooney | medium term just make creation on the fly work with the fallback to reading /sys for aviablie_isntances | |
| 14:06:39 | sean-k-mooney | in either case this is next week us problem i assume | |
| 14:06:44 | bauzas | agreed with the plan | |
| 14:07:05 | bauzas | I just need hands on deck | |
| 14:07:08 | sean-k-mooney | we proceed with teh current devstack plugin and enhance it to do the other bits as we go | |
| 14:07:16 | sean-k-mooney | *shall we | |
| 14:07:24 | dansmith | bauzas: I know you | |
| 14:07:36 | dansmith | are not suggesting that we're barking orders without helping right? | |
| 14:08:11 | bauzas | no no | |
| 14:08:31 | bauzas | this is just that this puzzle is horribly overcomplicated and I'm just a man | |
| 14:09:05 | bauzas | and now the situation has been properly addressed and acked by above, I get many sollicitations :) | |
| 14:09:13 | bauzas | (which is eventually good) | |
| 14:09:31 | bauzas | so, yeah, taking one piece after the other | |
| 14:13:05 | sean-k-mooney | dansmith: oh you found out why the kvm kerenl fialed. ya if mdev was compiled out in that it makes total sense | |
| 14:13:43 | opendevreview | Pavlo Shchelokovskyy proposed openstack/nova master: Forbid use_cow_images together with flat images_type https://review.opendev.org/c/openstack/nova/+/898229 | |
| 14:13:46 | sean-k-mooney | im pulling th trigger on the devstack plugin patch lets impove it in followups | |
| 14:15:56 | dansmith | sean-k-mooney: yep | |
| 16:25:21 | opendevreview | Merged openstack/nova master: Compile mdev samples for nova-next https://review.opendev.org/c/openstack/nova/+/897708 | |
| #openstack-nova - 2023-10-14 | |||
| 00:40:33 | opendevreview | Merged openstack/nova master: Warn if we find compute services in cell0 https://review.opendev.org/c/openstack/nova/+/897246 | |
| 03:20:24 | opendevreview | Takashi Natsume proposed openstack/nova master: Update contributor guide for 2024.1 Caracal https://review.opendev.org/c/openstack/nova/+/896305 | |
| #openstack-nova - 2023-10-16 | |||
| 11:27:34 | opendevreview | HanGuangyu proposed openstack/nova master: Unify ServersController._flavor_id_from_req_data param to server_dict https://review.opendev.org/c/openstack/nova/+/898315 | |
| 12:09:54 | ioni | hi guys | |
| 12:10:47 | ioni | i have an issue with deleting an instance after upgrading to Zed(I think). The error I'm having is: Requested operation is not valid: cannot undefine domain with nvram | |
| 12:10:50 | ioni | https://paste.xinu.at/qvp4dm/ | |
| 12:11:14 | ioni | the only thing that I changed is to add hw_machine_type=q35 to that specific image | |
| 12:11:23 | ioni | and is kinda random | |
| 12:11:51 | ioni | to delete the instance successfully i have to run virsh undefine domain --nvram first | |
| 12:12:25 | ioni | the problems occurs randomly on instances that sometimes are newly built with that image and deleted after | |
| 12:22:48 | ioni | i think i need to add to the image hw_firmware_type uefi for nova to understand to undefine nvram first before undefining the domain | |
| 13:17:56 | opendevreview | ribaudr proposed openstack/nova-specs master: Re-propose "Allow Manila shares to be directly attached to an instance when using libvirt" for Caracal https://review.opendev.org/c/openstack/nova-specs/+/898319 | |
| 13:20:12 | opendevreview | ribaudr proposed openstack/nova-specs master: Re-propose "Allow Manila shares to be directly attached to an instance when using libvirt" for Caracal https://review.opendev.org/c/openstack/nova-specs/+/898319 | |
| 13:24:55 | opendevreview | ribaudr proposed openstack/nova-specs master: Re-propose "Allow local scaphandre directory to be mapped to an instance using virtiofs" for Caracal https://review.opendev.org/c/openstack/nova-specs/+/898320 | |
| 14:43:31 | dansmith | we've got something going on with the live migration jobs and maybe others | |
| 14:43:35 | dansmith | lots of post failures | |
| 14:46:44 | dansmith | "msg": "The `lxml` module is not importable. Check the requirements." | |
| 14:52:25 | dansmith | sean-k-mooney: do you know how we get ansible modules installed on the hosts? like the virt module, which has requirements for libvirt-python, etc | |
| 14:56:33 | opendevreview | Tobias Urdin proposed openstack/nova master: [WIP] Handle scaling of cputune.shares https://review.opendev.org/c/openstack/nova/+/898326 | |
| 14:59:09 | opendevreview | Tobias Urdin proposed openstack/nova master: [WIP] Handle scaling of cputune.shares https://review.opendev.org/c/openstack/nova/+/898326 | |
| 14:59:46 | tobias-urdin | sean-k-mooney[m]: feedback on ^ – doesn't handle all cases yet but i dont want to spend time on it if it's just garbage then we'll just carry something similar downstream | |
| 14:59:46 | tobias-urdin | sean-k-mooney[m]: feedback on ^ – doesn't handle all cases yet but i dont want to spend time on it if it's just garbage then we'll just carry something similar downstream | |
| 15:08:14 | dansmith | sean-k-mooney: https://6df0d90b08f38024b3b0-ddacc1cdacca3582cad1fbbc53a5babe.ssl.cf2.rackcdn.com/897674/1/check/nova-live-migration/beaa1e0/job-output.txt | |
| 15:10:10 | sean-k-mooney | dansmith: so my guess is we need to set the interperter path for the remote node | |
| 15:11:24 | dansmith | sean-k-mooney: to what? something inside our new venv? | |
| 15:11:36 | sean-k-mooney | to the python3 binary in the venv | |
| 15:11:59 | dansmith | still, something changed over the weekend and I don't know what that is.. maybe the ansible module got changed or libvirt-python changed to a soft req on lxml or something? | |
| 15:12:38 | sean-k-mooney | ya did any changes happen in devstack we can look at this after the meeting | |
| 15:13:24 | dansmith | not really seeing anything | |
| 15:33:38 | opendevreview | Dan Smith proposed openstack/nova master: Install lxml before we need it in post-run https://review.opendev.org/c/openstack/nova/+/898435 | |
| 16:34:41 | dansmith | sean-k-mooney: bauzas this ^ seems to be working, it already passed the live-migration-ceph job | |
| 16:35:01 | sean-k-mooney | sweet i have it open but had not refreshed | |
| 16:35:08 | sean-k-mooney | ah i assume your watching the live output | |
| 16:35:55 | dansmith | yup | |
| 16:36:09 | dansmith | just passed the main job as well | |
| 16:39:02 | sean-k-mooney | added link to the passing job where it ran and left a +2 | |
| 16:39:48 | sean-k-mooney | if we start seeing this type of dep issue often i woudl prefer to use bindep or similar to store the packages in a file or move the list of packagges to a role default | |
| 16:39:54 | sean-k-mooney | but for now this is fine | |
| 16:41:42 | dansmith | ack, well, I'd like to get the gate unblocked at least | |
| 16:42:26 | dansmith | I know bindep is what we use for that stuff normally, but I figured it was more aimed at setup for running the services and not just for ansible stuff to work | |
| 16:42:34 | dansmith | I suppose it's also usable | |
| 16:42:45 | sean-k-mooney | bindep has tags | |
| 16:42:51 | sean-k-mooney | so we can use [ansible] | |
| 16:43:02 | sean-k-mooney | for example if we wanted to store ansible only deps or whatever | |
| 16:43:22 | dansmith | ack | |