| opendevreview | Jie Song proposed openstack/nova master: hardware: Fix image_meta.id within get_mem_encryption_constraint https://review.opendev.org/c/openstack/nova/+/904197 | 04:26 |
|---|---|---|
| opendevreview | Takashi Kajinami proposed openstack/nova master: GMR: Omit signal handleris in wsgi app https://review.opendev.org/c/openstack/nova/+/906279 | 05:11 |
| opendevreview | Takashi Kajinami proposed openstack/nova master: GMR: Omit signal handleris in wsgi app https://review.opendev.org/c/openstack/nova/+/906279 | 05:12 |
| opendevreview | Takashi Kajinami proposed openstack/nova master: GMR: Omit signal handlers in wsgi app https://review.opendev.org/c/openstack/nova/+/906279 | 05:54 |
| opendevreview | Jie Song proposed openstack/nova master: hardware: Fix image_meta.id within get_mem_encryption_constraint https://review.opendev.org/c/openstack/nova/+/904197 | 07:01 |
| TK__ | Hello Team,I have an openstack environment with 4 compute Nodes however yesterday out of the blue, the Instances on one of the compute nodes started shutting themselves down. Then I realized the the compute service was disabled when I checked in Horizon. I enabled it and restarted the instances and all was well but I cant figure out what caused it. I have attached the nova_compute logs here incase someone would like to support | 08:05 |
| TK__ | https://paste.openstack.org/show/bxqpgjvqT3DlOYWoeejc/ | 08:05 |
| bauzas | TK__: 2024-01-23 08:27:49.761 6 INFO nova.compute.manager [req-d9713a02-2cfd-4503-9dd8-cce1a0db0c73 - - - - -] [instance: 6d26e7bc-363a-4d0d-90c4-8708332f9621] Destroying instance as it has been evacuated from this host but still exists in the hypervisor | 08:28 |
| bauzas | TK__: looks like your compute somehow failed and you issued evacuations that were in progress | 08:28 |
| bauzas | once the compute went back up, nova started to delete the rogue evacuated instances | 08:29 |
| TK__ | We actually didnt evacuate them.... | 08:33 |
| *** LarsErik1 is now known as LarsErikP | 08:35 | |
| bauzas | TK__: you could try to look at the instance actions API | 08:35 |
| bauzas | pick one instance UUID that's been shutdown and see all its actions | 08:36 |
| TK__ | I see it evacuated but I think it was done by Masakari but not a user | 08:41 |
| bauzas | doh, I was about to reply | 08:47 |
| bauzas | TK_: (if you still read the logs) you can find the user/project IDs that issued the evacuations in the instance action API return | 08:48 |
| bauzas | but tbc, if you're using some top-level API that manages instance HA, you should rather look at their logs to see what happened and you shouldn't IMHO force to restart the compute instances, if they were evacuated | 08:49 |
| opendevreview | Jie Song proposed openstack/nova master: hardware: Fix image_meta.id within get_mem_encryption_constraint https://review.opendev.org/c/openstack/nova/+/904197 | 09:46 |
| pslestang | Hello team, do you know if there is a will to make this WIP (https://review.opendev.org/c/openstack/nova/+/508345) work in cinderclient part of nova? We'd like to get it working to improve requests reliabily and use options such as connect_retries, connect_retry_delay and so on. | 10:10 |
| Roamer` | sean-k-mooney, sorry we are only asking this after your zswap patches have been reviewed and merged (and thanks for your work BTW), but, hm, shouldn't configure_zswap() check for ENABLE_ZSWAP and not for ENABLE_KSMTUNED? I mean at https://review.opendev.org/c/openstack/devstack/+/890693/12/lib/host#40 | 10:16 |
| Roamer` | (FTR, this made the StorPool Cinder CI jobs fail because the z3fold kernel module is not loaded by default on Ubuntu Jammy... which might be a different problem related to the way we create our own images, we do not use nodepool-builder; we shall find a way to work around that, but it was kind of surprising that the Cinder jobs tried to enable zswap at all) | 10:18 |
| frickler | Roamer`: oh, do you happen to be responsible for/related to the StorPool CI? we would really really like it to stop voting on devstack and other unrelated projects, where it only produces garbage. we were pretty close to actually disabling the account for it even | 10:21 |
| Roamer` | frickler, wait, it is trying to vote?! argh | 10:22 |
| Roamer` | sorry about that! | 10:22 |
| Roamer` | yeah, okay, we will look into it ASAP | 10:22 |
| bauzas | pslestang: that patch is very old, and we changed our opinions in between | 10:22 |
| frickler | Roamer`: see for example the devstack patch you just mentioned | 10:23 |
| bauzas | pslestang: now, we'd prefer to now use sdk instead of cinderclient | 10:23 |
| bauzas | pslestang: https://bugs.launchpad.net/nova/+bug/1996210 | 10:23 |
| pslestang | bauzas: ok understood thx | 10:30 |
| sean-k-mooney | Roamer`: yes it should | 10:48 |
| sean-k-mooney | Roamer`: i can put up a patch to fix that quickly its just a silly copy past error | 10:49 |
| sean-k-mooney | i have been testing this with everything turned on on 22.04 ubuntu cloud images however so im surpised it cause you issues | 10:50 |
| sean-k-mooney | i also spot checked it on centos9 stream to test the rpm based distros | 10:51 |
| sean-k-mooney | Roamer`: https://review.opendev.org/c/openstack/devstack/+/906504 that should fix it. until that is merged you can add ENABLE_KSMTUNED=false as a workaround | 10:55 |
| sean-k-mooney | frickler: dansmith ^ if ye have time for that fix please take a look | 11:00 |
| frickler | sean-k-mooney: +3 already | 11:00 |
| sean-k-mooney | :) thanks | 11:01 |
| bauzas | gosh the scheduler manager UTs framework is sooo crazy, I'm buggy from the last hours | 12:48 |
| *** blarnath is now known as d34dh0r53 | 15:17 | |
| auniyal | bauzas, gibi I understand all conf will be updated under nova/conf dir | 15:29 |
| auniyal | but is there a way we can update update this page as well https://docs.openstack.org/ocata/config-reference/compute/config-options.html | 15:29 |
| bauzas | the conf doc page is automatically generated | 15:29 |
| bauzas | you just don't have the right link :) | 15:30 |
| auniyal | this is the first page got when I googled for conf option | 15:30 |
| bauzas | https://docs.openstack.org/nova/latest/configuration/config.html | 15:30 |
| auniyal | yeah, this one I am aware of, and I understand whatever conf opt we want to update, update under nova/conf, but the link I shared is the first link comes up when we google nova conf variables | 15:33 |
| auniyal | so I think this should be updated too | 15:33 |
| auniyal | or is someone else maintaing it | 15:33 |
| bauzas | auniyal: the link you refer is some Ocata release doc | 15:39 |
| bauzas | so that's an EOL release | 15:40 |
| auniyal | yeah, and its not opening for zed too, but it has nice table format | 15:40 |
| bauzas | you can get docs from EOL releases for sure (as people can still operate them) but people looking at docs should target the right release docs | 15:41 |
| bauzas | auniyal: because we changed the docs structure somewhere around Ussuri-ish | 15:41 |
| auniyal | it must be just template right | 15:41 |
| auniyal | okay, so updating like that may be lot of backport | 15:41 |
| auniyal | shall we do it ? i think its more informative/user friendly then https://docs.openstack.org/nova/latest/configuration/config.html | 15:43 |
| bauzas | auniyal: you don't understand, we shouldn't touch the ocata documentation *at all* | 16:05 |
| bauzas | if we were touching some stable docs, this would be due to some config backport we have in our code | 16:05 |
| bauzas | but given ocata is EOL, this is now pristine doc | 16:05 |
| auniyal | ack thanks bauzas for clearing it | 16:06 |
| bauzas | that's not because we're bad at Google's SEO that we should temper our docs | 16:06 |
| opendevreview | Amit Uniyal proposed openstack/nova master: enforce remote console shutdown https://review.opendev.org/c/openstack/nova/+/901824 | 16:37 |
| gibi | bauzas: sean-k-mooney: the master revert of https://review.opendev.org/q/topic:%22power-mgmt-fixups%22 is finally grean | 17:33 |
| gibi | green even | 17:33 |
| bauzas | +2d | 17:35 |
| bauzas | I can mark the series as a prio in the etherpad | 17:35 |
| bauzas | (done) | 17:36 |
| gibi | thanks | 17:38 |
| opendevreview | Sylvain Bauza proposed openstack/nova master: Fix verifying all the alloc requests from a multi-create https://review.opendev.org/c/openstack/nova/+/846786 | 18:26 |
| opendevreview | sean mooney proposed openstack/nova master: add initial healthcheck support https://review.opendev.org/c/openstack/nova/+/825015 | 19:43 |
| opendevreview | sean mooney proposed openstack/nova master: add healthcheck manager to manager base https://review.opendev.org/c/openstack/nova/+/827844 | 19:43 |
| opendevreview | sean mooney proposed openstack/nova master: add healthcheck tracker to nova context https://review.opendev.org/c/openstack/nova/+/829468 | 19:43 |
| opendevreview | sean mooney proposed openstack/nova master: add healthcheck utils and constants https://review.opendev.org/c/openstack/nova/+/829469 | 19:43 |
| opendevreview | sean mooney proposed openstack/nova master: add healthcheck endpoint to proxy commands https://review.opendev.org/c/openstack/nova/+/830703 | 19:43 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!