| *** mhen_ is now known as mhen | 01:54 | |
| ralonsoh | frickler, now the patch https://review.opendev.org/c/openstack/devstack/+/955275 is requesting my to remove the definition in 2023.1 | 05:49 |
|---|---|---|
| ralonsoh | but I have this problem: https://review.opendev.org/c/openstack/devstack/+/955338 | 05:50 |
| ralonsoh | --> The nodeset "openstack-two-node-jammy" was not found. | 05:50 |
| ralonsoh | But I can't/shouldn't remove this nodeset from unmantained | 05:50 |
| frickler | ralonsoh: don't get confused by software-factory. the real zuul wants you to drop xenial and fedora. but this is also another good example of how much load the unmaintained branches generate for people who don't really care about them | 06:21 |
| ralonsoh | frickler, sorry, I don't get that | 06:37 |
| ralonsoh | xenial and fedora? | 06:38 |
| frickler | ralonsoh: the first comment on https://review.opendev.org/c/openstack/devstack/+/955338 , ubuntu-xenial and fedora-36 to be more precise | 07:12 |
| ralonsoh | frickler, but we should not remove these nodepools from 2023.1 | 07:18 |
| ralonsoh | well, yes | 07:18 |
| ralonsoh | xenial was supported up to 2021 | 07:19 |
| ralonsoh | and fedora36 may 16, 2023 | 07:19 |
| ralonsoh | should I remove them too? | 07:19 |
| frickler | yes, these are no longer available, since nobody wrote code to build them with zuul-launcher | 07:21 |
| ralonsoh | perfect, thanks! | 07:21 |
| ralonsoh | Now up to victoria | 07:35 |
| ralonsoh | this is going to be a nightmare | 07:35 |
| ralonsoh | victoria patch: https://review.opendev.org/c/openstack/devstack/+/955343 | 07:43 |
| frickler | looks like gerritbot is taking a break, I'll check that in a bit. | 08:34 |
| opendevreview | Rodolfo Alonso proposed openstack/devstack master: [Neutron] Add TARGET_ENABLE_OVN_AGENT variable to enable OVN agent https://review.opendev.org/c/openstack/devstack/+/955187 | 08:35 |
| frickler | ah, no, it just thankfully ignores unmaintained branches | 08:36 |
| opendevreview | Rodolfo Alonso proposed openstack/devstack master: [Neutron] Create method to start/stop OVN agents https://review.opendev.org/c/openstack/devstack/+/955348 | 08:53 |
| sean-k-mooney | ralonsoh: im surpised you still have unmatined branches older then victora or even victoria | 08:55 |
| ralonsoh | sean-k-mooney, no we don't | 08:55 |
| sean-k-mooney | elodilles cares about victoria to some degree | 08:55 |
| sean-k-mooney | oh ok | 08:55 |
| ralonsoh | but I'm trying to push a patch for devstack | 08:55 |
| ralonsoh | and the dependencies go down to victoria | 08:56 |
| ralonsoh | https://review.opendev.org/q/topic:%22remove_opensuse-15%22 | 08:56 |
| sean-k-mooney | oh dropping suse | 08:56 |
| ralonsoh | that was caused by https://review.opendev.org/c/openstack/project-config/+/955214 | 08:56 |
| ralonsoh | yes | 08:56 |
| sean-k-mooney | im confused how that is merged on thabel but not unmaintined or master but i probaly dont want to knwo why | 08:57 |
| sean-k-mooney | s/thable/stable/ | 08:59 |
| ralonsoh | today I'm having a funny day with devstack and grenade. I found we are NOT restarting the OVN agents in grenade... | 09:00 |
| sean-k-mooney | well in theory you dont need too | 09:01 |
| sean-k-mooney | because you are not changing the os version and therfor not changing the package | 09:01 |
| ralonsoh | no no | 09:01 |
| sean-k-mooney | although you shoudl be restreating the neutron ones | 09:01 |
| ralonsoh | not the ovn-controller and backend things | 09:01 |
| ralonsoh | but OVN Metadata or OVN agent | 09:01 |
| sean-k-mooney | oh you mena the ovn metadata agent right | 09:01 |
| ralonsoh | yes | 09:02 |
| sean-k-mooney | ya that woudl cuase interesting... bugs | 09:02 |
| ralonsoh | yeah... don't tell me, I'm a bit scared right now | 09:02 |
| ralonsoh | https://review.opendev.org/c/openstack/neutron/+/955349 | 09:02 |
| sean-k-mooney | you shoudlnt be sared, worst case you jsut getting extra testing of your backward compat mult nod grended intentioally only runs greande on one of the 2 nodes so that the other node continues to run the old code | 09:04 |
| sean-k-mooney | what is concernign is obvioulsy no testign with themix | 09:05 |
| sean-k-mooney | and only testing with the old version of the metadta and new version of everything else | 09:05 |
| ralonsoh | yeah, that's right: old metadata code is working with new API/backend | 09:05 |
| frickler | ralonsoh: don't invest much time into unmaintained branches unless you really want to, we can just force-merge 955343 etc. and let elodilles_pto deal with the fallout if needed | 10:02 |
| ralonsoh | frickler, I can try removing the failing job openstacksdk-functional-devstack | 10:03 |
| ralonsoh | then I'll continue pushing patches for newer versions | 10:03 |
| frickler | ralonsoh: commented on two patches where you missed a nodeset. if you update these I can force-merge the whole stack, I don't have normal core rights on these anyway, but we need to unblock the master branch | 10:29 |
| ralonsoh | frickler, pushing new PS right now | 10:36 |
| opendevreview | Tanvi Nautiyal proposed openstack/tempest master: Improve Swift Quota test with exceeding upload https://review.opendev.org/c/openstack/tempest/+/954354 | 11:45 |
| frickler | ralonsoh: merged, seems we're still missing yoga+zed, do you want to propose those, too? | 12:00 |
| ralonsoh | frickler, yes, right now | 12:35 |
| ralonsoh | yoga: https://review.opendev.org/c/openstack/devstack/+/955368 | 12:40 |
| ralonsoh | zed: https://review.opendev.org/c/openstack/devstack/+/955369 | 12:43 |
| frickler | merged, now tempest is the next blocker and it looks like gmaan will have some work to do to clean this up | 13:01 |
| frickler | I wonder if instead of dropping the devstack-single-node-opensuse-15 nodeset we could just replace the label with something else for now, like debian-bookworm. cc fungi clarkb | 13:02 |
| opendevreview | Rodolfo Alonso proposed openstack/tempest master: Remove job ``tempest-full-py3-opensuse15`` https://review.opendev.org/c/openstack/tempest/+/955375 | 13:35 |
| ralonsoh | frickler, ^ | 13:35 |
| ralonsoh | ah, I didn't see your messages | 13:35 |
| ralonsoh | ufff now the tempest test is blocked by glace in victoria | 13:37 |
| frickler | yes, that's going to be a neverending story | 13:42 |
| opendevreview | Eric Harney proposed openstack/devstack-plugin-nfs master: WIP: Get nfs-fips job running https://review.opendev.org/c/openstack/devstack-plugin-nfs/+/955317 | 13:50 |
| clarkb | frickler: yes, one hack to make a nodeset keep working in zuul config validation is to set the label to something else | 14:43 |
| clarkb | frickler: however, didn't we remove opensuse 15 a while back? I don't know if we need to do anything special at this point to make it "work" | 14:43 |
| frickler | clarkb: well random changes touching .zuul.yaml are failing, like e.g. https://review.opendev.org/c/openstack/devstack/+/953406 and the cleanup https://review.opendev.org/c/openstack/devstack/+/955275 is blocked by tempest | 14:53 |
| frickler | and the cleanup for tempest is again blocked by lots of other unmaintained branches | 14:54 |
| clarkb | I see tempest uses the nodeset so we can't trivially remove the nodeset. In that case then yes updating that nodeset's label to point at something else (or even be an empty label might work) | 14:58 |
| clarkb | or we would have to force merge udpates to devstack (and maybe tempest) | 14:58 |
| clarkb | doing the label hack is probably a reasonable workaround | 14:58 |
| opendevreview | Dr. Jens Harbott proposed openstack/devstack master: Remove the nodeset "devstack-single-node-opensuse-15" https://review.opendev.org/c/openstack/devstack/+/955275 | 15:16 |
| opendevreview | Dr. Jens Harbott proposed openstack/devstack master: Fix the nodeset "devstack-single-node-opensuse-15" https://review.opendev.org/c/openstack/devstack/+/955275 | 15:19 |
| frickler | clarkb: ^^ a nodeset without nodes isn't allowed, how would an empty label look like? for now I just replaced opensuse with bookworm | 15:20 |
| clarkb | frickler: I think you have to do nodes: [] | 15:21 |
| clarkb | but the approach in the last patchset should work so +2 from me | 15:22 |
| frickler | doh, empty list, sure. let me try that first, would seem less confusing | 15:23 |
| opendevreview | Dr. Jens Harbott proposed openstack/devstack master: Fix the nodeset "devstack-single-node-opensuse-15" https://review.opendev.org/c/openstack/devstack/+/955275 | 15:25 |
| opendevreview | Dr. Jens Harbott proposed openstack/devstack master: Fix the nodeset "devstack-single-node-opensuse-15" https://review.opendev.org/c/openstack/devstack/+/955275 | 15:25 |
| *** haleyb is now known as haleyb|out | 18:33 | |
| opendevreview | Merged openstack/devstack master: Fix the nodeset "devstack-single-node-opensuse-15" https://review.opendev.org/c/openstack/devstack/+/955275 | 19:52 |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.org will be offline briefly for a configuration and version update, but should return to service momentarily | 20:07 | |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!