| *** jpena|off is now known as jpena | 07:23 | |
| lajoskatona | kopecmartin: Hi, I opened a bug against Neutron/Tempest and Nova: https://bugs.launchpad.net/neutron/+bug/2017023 | 07:32 |
|---|---|---|
| lajoskatona | kopecmartin: this is to track the old network APIs in Nova question | 07:32 |
| lajoskatona | kopecmartin: I plan to check the coverage of such tests under tempest/api/compute to compare with tempest/api/network and tempest/scenario | 07:33 |
| lajoskatona | kopecmartin: please check it if you have some free time, and I am happy to join QA meeting or similar if you think it is necessary to discuss and see all aspects of this topic | 07:34 |
| dansmith | we're good to drop any py38 unit and functional jobs from master now right? | 14:11 |
| kopecmartin | lajoskatona: thanks, i was thinking about that recently, i put it to qa-priority goals so that i find time for that | 14:35 |
| kopecmartin | lajoskatona: or it's not the same thing as this https://etherpad.opendev.org/p/neutron-tempest-test_duplications | 14:37 |
| lajoskatona | kopecmartin: thanks, on nova channel there was some comments, I add those with links to the lp bug | 14:37 |
| lajoskatona | kopecmartin: no, that is just Neutron tests between tempest and neutron-tempest-plugin | 14:37 |
| kopecmartin | right | 14:38 |
| opendevreview | Ihar Hrachyshka proposed openstack/devstack master: git: support git checkout for a commit hash https://review.opendev.org/c/openstack/devstack/+/880892 | 15:08 |
| *** jpena is now known as jpena|off | 16:17 | |
| gmann | kopecmartin: lajos moving to neutron API is good but nova API should keep working, do you know where they are broken ? | 16:28 |
| gmann | dansmith: yes, py38 is not needed on master. unit test jobs are dropped by common template unless projects explicitly adding them. | 16:29 |
| dansmith | gmann: ack, we still have functional jobs on 38 in glance, some of which are failing now because castellan has required >=3.9 apparently | 16:30 |
| dansmith | so just want to make sure we're good to bump those, but found a patch from rosmaita already proposed so that's cool | 16:30 |
| dansmith | however, there are functional fips jobs still pinned to 3.8 apparently, so not sure how we're going to work around those | 16:30 |
| gmann | kopecmartin: lajos, commented on bug to keep those tests https://bugs.launchpad.net/neutron/+bug/2017023 | 16:33 |
| gmann | dansmith: ok, yeah we need to move those to py39. and slowly projects will be making in their setup.cfg >=3.9 then py3.8 will be not testable also | 16:34 |
| dansmith | yeah, apparently castellan has already done that | 16:34 |
| clarkb | is grenade done running on focal on the old side? (focal is 3.8) | 16:35 |
| dansmith | no, grenade is still on focal AFAIK | 16:35 |
| gmann | I think it is on jammy as 2023.1 was testing jammy as main distro | 16:38 |
| gmann | yes on master grenade (2023.1->master) running on jammy https://github.com/openstack/grenade/blob/d8294440e7dede370b5ee80facf49a511dc7eb09/.zuul.yaml#L163 | 16:39 |
| dansmith | is it? I thought it was still focal | 16:39 |
| gmann | skip level is on focal | 16:39 |
| dansmith | ah right right, I'm thinking of skip level | 16:39 |
| gmann | oh no, skip level also on jammy and we said let's see how long it is ok on jammy | 16:40 |
| gmann | because focal is not supported in master so running on focal is issue | 16:40 |
| gmann | best chance was trying zed on jammy -> master upgrade | 16:40 |
| gmann | *skip-level-always job | 16:41 |
| dansmith | meaning zed->master? | 16:43 |
| dansmith | localhost | Label: ubuntu-jammy | 16:43 |
| gmann | yeah | 16:43 |
| dansmith | jammy wasn't official on zed right? maybe we just said "it happens to work on jammy" ? | 16:44 |
| gmann | yeah, that is best bet we discussed | 16:44 |
| gmann | because there are more chance to have master failing on focal than zed on jammy | 16:45 |
| dansmith | yeah okay I guess that's why I thought it was still focal | 16:45 |
| dansmith | because yeah: https://governance.openstack.org/tc/reference/runtimes/zed.html | 16:45 |
| dansmith | and I guess grenade was still on focal for 2023.1 | 16:45 |
| dansmith | so this is the problem for us: https://opendev.org/openstack/openstack-zuul-jobs/src/branch/master/zuul.d/jobs.yaml#L62-L66 | 16:47 |
| gmann | dansmith: grenade on stable/2023.1 which does zed->2023.1 is on focal | 16:47 |
| dansmith | fips still runningon focal | 16:47 |
| dansmith | oh, no, | 16:48 |
| dansmith | there's a jammy above | 16:48 |
| gmann | yeah it is with branch variant | 16:48 |
| dansmith | right okay lemme try to put up a patch | 16:48 |
| dansmith | rosmaita's patch to glance said we couldn't convert yet, but maybe he was mis-reading that | 16:48 |
| gmann | AFAIK i think ceph job is still running on focal therwise all are on jammy on master | 16:49 |
| gmann | k | 16:49 |
| dansmith | ooh | 16:49 |
| dansmith | because of the cephadm stuff? | 16:49 |
| gmann | yeah | 16:49 |
| gmann | this is still failing https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/865315 | 16:50 |
| gmann | which tried to move to Jammy | 16:50 |
| dansmith | yeah | 16:51 |
| dansmith | gmann: so I don't see a 310-fips job (or 39 for that matter) | 16:51 |
| dansmith | https://opendev.org/openstack/openstack-zuul-jobs/src/branch/master/zuul.d/jobs.yaml#L865 | 16:51 |
| clarkb | with fips jammy didn't have the kernel available yet. Focal does but then iscsi uses md5 inappropriately | 16:51 |
| dansmith | ah | 16:52 |
| clarkb | I think the idea was to pause fips on ubuntu until jammy got the kernel stuff | 16:52 |
| dansmith | the above job is centos-based, fwiw | 16:52 |
| dansmith | centos8 that is | 16:52 |
| gmann | ah right. that is also on focal. | 16:53 |
| gmann | ade_lee: might know the plan | 16:53 |
| ade_lee | gmann, dansmith yup - the kernel stuff needed for iscsi is in jammy, but fips support is not available for jammy | 16:56 |
| ade_lee | so we can use ubuntu focal or fips only if you don't need iscsi | 16:57 |
| ade_lee | like for ironic perhaps | 16:57 |
| ade_lee | jammy fips is supposed to be available sometime next year. | 16:58 |
| ade_lee | till then, we have centos - and if need be for stability, rocky | 16:58 |
| ade_lee | gmann, btw - I responded to the comment on the review -- I can change to checking enable_fips if you prefer | 16:59 |
| dansmith | ade_lee: yeah problem is some projects have already put python>=3.9 in their requirements, which means any project/job that requires that other package can't run on focal anymore | 16:59 |
| ade_lee | dansmith, yeah - so no dice on ubuntu till next year then | 17:01 |
| ade_lee | dansmith, when enabling fips, the subscription agent just errors out on jammy. I can check with canonical to see if there is any other option | 17:02 |
| dansmith | ack, I'm just saying we've got a potentially conflicting set of requirements coming here :) | 17:03 |
| gmann | yeah those will start breaking fips job | 17:03 |
| gmann | ade_lee: thanks, I will the comment today | 17:03 |
| ade_lee | yup - and this might be a problem when we get fips on jammy, because by then we'll be using python>=3.10 or whatnot .. | 17:04 |
| dansmith | jammy is 3.10 | 17:05 |
| gmann | ade_lee: for current master 2023.2 it can be >-py3.9 and in future if we bump the python min version to py310 then it can be >=py3.10 | 17:05 |
| gmann | but py3.9 is min version supported a it is in debian 11 | 17:05 |
| ade_lee | cool | 17:06 |
| gmann | https://governance.openstack.org/tc/reference/runtimes/2023.2.html#python-runtimes-for-2023-2 | 17:06 |
| opendevreview | Ghanshyam proposed openstack/tempest master: DNM: testing volume attachments https://review.opendev.org/c/openstack/tempest/+/880891 | 17:16 |
| opendevreview | Ghanshyam proposed openstack/tempest master: DNM: testing volume attachment https://review.opendev.org/c/openstack/tempest/+/881132 | 18:08 |
| opendevreview | Ghanshyam proposed openstack/tempest master: DNM: testing volume attachment https://review.opendev.org/c/openstack/tempest/+/881132 | 19:41 |
| opendevreview | Ghanshyam proposed openstack/tempest master: DNM: testing volume attachment https://review.opendev.org/c/openstack/tempest/+/881132 | 21:20 |
| gmann | ade_lee: agree with your point in FIPs change, +A | 22:19 |
| gmann | https://review.opendev.org/c/openstack/devstack/+/871606 | 22:19 |
| opendevreview | Merged openstack/devstack master: Modify devstack-base to allow for fips https://review.opendev.org/c/openstack/devstack/+/871606 | 23:55 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!