| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2023-10-25 | |||
| 08:05:14 | opendevreview | Amit Uniyal proposed openstack/nova stable/2023.1: Translate VF network capabilities to port binding https://review.opendev.org/c/openstack/nova/+/898945 | |
| 08:05:50 | opendevreview | Amit Uniyal proposed openstack/nova stable/zed: Translate VF network capabilities to port binding https://review.opendev.org/c/openstack/nova/+/899229 | |
| 09:04:26 | opendevreview | Sylvain Bauza proposed openstack/nova master: Fix rebuild compute RPC API exception for rolling-upgrades https://review.opendev.org/c/openstack/nova/+/899235 | |
| 09:04:26 | opendevreview | Sylvain Bauza proposed openstack/nova master: add a regression test for all compute RPCAPI 6.x pinnings for rebuild https://review.opendev.org/c/openstack/nova/+/899234 | |
| 09:04:41 | bauzas | noonedeadpunk: ^ | |
| 09:05:21 | bauzas | I'll look at how grenade can verify rebuild instances | |
| 09:07:41 | noonedeadpunk | bauzas: thanks for working on that! | |
| 09:08:07 | bauzas | once it's approved, I'll backport those down to Zed | |
| 09:16:15 | bauzas | interesting, I only find 2 tempest tests that verify rebuild | |
| 09:16:45 | bauzas | https://github.com/openstack/tempest/blob/d1742bad94e94ebabe875ed24415d1ff510a5ae7/tempest/api/compute/servers/test_server_actions.py#L395-L473 | |
| 09:40:31 | gibi | kashyap: thanks for the patch please keep me reminded about it if I don't show up reviewing it. This week and next week is pretty hectic | |
| 09:40:55 | gibi | bauzas: maybe we need to do more public shaming of people to get them to triage bugs :D | |
| 09:41:16 | kashyap | gibi: Right, it'll be the week after. The patch doesn't deserve your review yet, but I'll need some help as I want to be careful not to introduce any other subtle issues | |
| 09:42:17 | gibi | kashyap: I think you can create a local reproduction for it with a multi node devstack if you set the cpu_mode to host model (and maybe tweek the devstack VMs so libvirt reports different cpu for capabilities and domcapabiltities) | |
| 09:42:51 | gibi | or you can try the same in an upstream CI patch | |
| 09:43:15 | gibi | in the CI we set cpu_mode=custom so you can try to change that to host-model | |
| 09:43:26 | kashyap | gibi: The DevStack VMs (level-1 guests) should be CentOS9 stream, yeah? | |
| 09:43:41 | gibi | kashyap: in my case the lvl1 was centos stream 9 | |
| 09:43:56 | gibi | but I think for the repro it is enough the have different cap and domcap reported | |
| 09:44:00 | gibi | + host-model | |
| 09:44:08 | kashyap | Yeah | |
| 09:44:15 | kashyap | gibi: For an upstream CI test, which job do you suggest to change? | |
| 09:45:23 | gibi | we have a live migration job | |
| 09:45:57 | gibi | (ahh my other devstack based on ubuntu lvl1 reports the same for cap and domcap so lvl1 OS matters) | |
| 09:46:36 | gibi | kashyap: this is our live migration job https://github.com/openstack/nova/blob/master/.zuul.yaml#L102 | |
| 09:46:37 | kashyap | Ah, good that you checked | |
| 09:47:09 | kashyap | (On Ubuntu setup) | |
| 09:47:17 | gibi | you can try to add | |
| 09:47:18 | gibi | $NOVA_CPU_CONF: | |
| 09:47:18 | gibi | post-config: | |
| 09:47:18 | gibi | devstack_local_conf: | |
| 09:47:32 | gibi | to that job | |
| 09:47:36 | gibi | to pass nova config | |
| 09:47:55 | gibi | similarly to https://github.com/openstack/nova/blob/master/.zuul.yaml#L400C1-L402C26 | |
| 09:51:19 | kashyap | gibi: Ah, thank you! Handy example | |
| 09:56:49 | bauzas | gibi: about the upstream tempest jobs, wdyt if I add https://github.com/openstack/tempest/blob/d1742bad94e94ebabe875ed24415d1ff510a5ae7/tempest/api/compute/servers/test_server_actions.py#L395-L473 to the grenade-multinode job ? | |
| 09:57:12 | bauzas | I don't see any other tempest test for rebuild apart from the volume one | |
| 09:57:45 | bauzas | in https://github.com/openstack/nova/blob/master/.zuul.yaml#L560 we only check the smoke tests and the live-migration ones | |
| 09:58:06 | bauzas | which is not a long list https://0fd77280b0334417b6ee-1d67ec5eb204f28641a7073258861323.ssl.cf5.rackcdn.com/877410/7/check/nova-grenade-multinode/4e93584/testr_results.html | |
| 09:58:41 | gibi | bauzas: yeah it make sense to add a rebuild test to grenade | |
| 09:58:52 | bauzas | oh wait, I'm wrong | |
| 09:58:54 | bauzas | https://github.com/openstack/tempest/blob/d1742bad94e94ebabe875ed24415d1ff510a5ae7/tempest/api/compute/servers/test_server_actions.py#L263 | |
| 09:59:21 | bauzas | oh no, we just test test_reboot_server_hard[id-2cb1baf6-ac8d-4429-bf0d-ba8a0ba53e32,smoke] | |
| 09:59:59 | bauzas | so, two possibilities : either we make rebuild_server a smoke test in tempest, or we change grenade | |
| 10:00:18 | gibi | there are simpler rebuild servers like https://github.com/openstack/tempest/blob/d1742bad94e94ebabe875ed24415d1ff510a5ae7/tempest/api/compute/servers/test_servers_microversions.py#L40 | |
| 10:01:11 | bauzas | test_rebuild_server seems enough simple to me https://github.com/openstack/tempest/blob/d1742bad94e94ebabe875ed24415d1ff510a5ae7/tempest/api/compute/servers/test_server_actions.py#L262-L278 | |
| 10:01:32 | bauzas | and it doesn't depend on microversion check | |
| 10:02:17 | gibi | OK, that is a good one. I think we can make that smoke | |
| 10:02:55 | bauzas | or we could add all the ServerActionsTestJSON tests to the grenade-multinode job | |
| 10:03:10 | gibi | I can be convinced to that too | |
| 10:03:12 | bauzas | you know what ? I'll quick ask the community today | |
| 10:03:20 | gibi | whaaat? how? :D | |
| 10:03:26 | bauzas | by the PTG :p | |
| 10:03:41 | gibi | OMG, you are so lucky we have a PTG :D | |
| 10:04:00 | gibi | we need more PTG :D lets make every week a PTG :D | |
| 10:04:12 | bauzas | lol | |
| 10:04:27 | bauzas | when I look at https://0fd77280b0334417b6ee-1d67ec5eb204f28641a7073258861323.ssl.cf5.rackcdn.com/877410/7/check/nova-grenade-multinode/4e93584/testr_results.html I think we're missing a lot of server actions | |
| 10:04:39 | kashyap | What did you have for breakfast, gibi? Share your tips; sharing is caring, your parents were right. | |
| 10:05:09 | gibi | I think it is combination of excercise + coffein :) | |
| 10:05:15 | kashyap | I'm on eucalyptus tea-only today due to sore throat, so low on energy. | |
| 10:05:27 | kashyap | gibi: Yes, that combo works | |
| 10:05:35 | gibi | ohh I have ginseng tea for the PTG part of the day ;) | |
| 10:06:10 | gibi | bauzas: yep, I think it make sense to increase the test scope of grenade | |
| 10:07:03 | bauzas | gibi: kashyap: ahah, for me I rushed to my local supermarket to just get coffee since today my machine was no longer having :;D | |
| 10:07:41 | kashyap | bauzas: Are supermarket (read-made) coffees any good? | |
| 10:08:07 | gibi | bauzas: you need more machines :D and a backup set of instant coffe :) | |
| 10:08:14 | kashyap | gibi: Thanks for the Ginseng idea, I should try it | |
| 10:10:02 | bauzas | kashyap: not that much, but I'm not a difficult guy | |
| 10:10:09 | bauzas | I can drink us coffee | |
| 10:10:20 | bauzas | so you can imagine | |
| 10:10:43 | kashyap | Yeah, I'm a slightly difficult guy with coffee. (And _very_ difficult guy with bread - I'm in the hunt for the "best sourdough bread" :D) | |
| 10:10:51 | gibi | I started to like US coffee as a warm soft drink. It has less coffein so I can drink a lot from it :) | |
| 10:10:57 | kashyap | (I know, I know, "best" is relative.) | |
| 10:11:40 | gibi | kashyap: ask sean-k-mooney[m] for a sourdough recipe | |
| 10:11:54 | kashyap | I didn't know sean-k-mooney[m] covered that too. Unfair! :D | |
| 10:13:43 | kashyap | I don't yet have an oven (I know); so I can't make it yet myself. Maybe I should search for a compact one | |
| 10:15:18 | gibi | sourdough is a journey as far as I heard. I bake simple cakes but not bread. | |
| 10:16:51 | kashyap | Yeah, I heard so too; for now I'm happy to buy it from a baker I know makes it well | |
| 10:21:02 | gibi | now I'm getting hungry :D | |
| 10:21:49 | kashyap | Go for it; it's lunchtime too :) | |
| 10:22:12 | gibi | true | |
| 10:51:33 | sean-k-mooney | i actully dont make sourdough because i dont like the taste but i make yested breads using similar techniques/recipies i just dont keep the starter an make a polish the night before instead | |
| 10:51:44 | sean-k-mooney | very similer texture without the sour taste | |
| 10:57:43 | gibi | interesting | |
| 11:17:54 | opendevreview | Amit Uniyal proposed openstack/nova stable/yoga: Fix the PCI device capability dict creation https://review.opendev.org/c/openstack/nova/+/899252 | |
| 11:17:55 | opendevreview | Amit Uniyal proposed openstack/nova stable/yoga: Store pf_mac_address and vf_num in extra_info https://review.opendev.org/c/openstack/nova/+/899253 | |
| 11:17:56 | opendevreview | Amit Uniyal proposed openstack/nova stable/yoga: Translate VF network capabilities to port binding https://review.opendev.org/c/openstack/nova/+/899254 | |
| 12:04:44 | whoami-rajat | bauzas, hey, just to confirm, the nova team will be joining the cinder room at 1600 UTC right? | |
| 12:06:58 | bauzas | whoami-rajat: YUUUUP https://etherpad.opendev.org/p/nova-caracal-ptg#L58 | |
| 12:08:57 | whoami-rajat | bauzas, great, thanks! | |
| 13:02:50 | bauzas | sean-k-mooney: joining ? | |
| 13:03:27 | bauzas | gmann: can you just us now ? | |
| 13:03:34 | bauzas | for SLURP discussions ? | |
| 13:03:39 | sean-k-mooney | yep | |
| 13:03:45 | sean-k-mooney | be there in a sec | |
| 13:03:48 | sean-k-mooney | feel free to start | |
| 13:05:38 | gmann | joining | |
| 13:27:33 | kashyap | bauzas: I'd appreciate if someone is taking notes of the PCI device scheduling thing. I'm on another call, but will join late to listen in | |
| 13:27:50 | sean-k-mooney | kashyap: we are not talking about that | |
| 13:28:01 | kashyap | Oh, okay; when will that topic be discussed? | |