| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2026-09-29 | |||
| 08:59:48 | gibi | ❯ reno new foo | |
| 08:59:50 | gibi | issues: List known issues here, or remove this section. All of the list items in | |
| 09:00:45 | gibi | we have no specific reno config in the repo as far as I see | |
| 09:00:47 | bauzas | yeah this is why I documented https://docs.openstack.org/nova/latest/contributor/releasenotes.html years before | |
| 09:01:07 | bauzas | lemme write a patch using issues and we'll see | |
| 09:01:19 | gibi | we have releasenotes/source/conf.py but not config.yaml | |
| 09:01:49 | gibi | bauzas: OK | |
| 09:02:37 | bauzas | I'll run the tox target on my laptop to check the doc | |
| 09:10:00 | bauzas | mmm, I don't see the section even after changing the section name :( | |
| 09:10:10 | bauzas | lemme push the patch and we'll see | |
| 09:10:45 | opendevreview | Sylvain Bauza proposed openstack/nova master: Fix graceful shutdown release note section https://review.opendev.org/c/openstack/nova/+/1007837 | |
| 09:12:11 | gibi | thanks | |
| 09:13:25 | opendevreview | Sylvain Bauza proposed openstack/nova master: Fix graceful shutdown release note section https://review.opendev.org/c/openstack/nova/+/1007837 | |
| 09:16:25 | bauzas | hmmmm | |
| 09:16:37 | bauzas | really weird, still doesn't show up | |
| 09:16:50 | bauzas | but I'm pretty sure that issues is the default name | |
| 09:16:57 | elodilles | i don't know the reason either :/ | |
| 09:18:13 | elodilles | the docs page refreshes like hourly, but i guess this is not the reason here | |
| 09:18:34 | bauzas | I'm not seeing it even on my laptop as generated page | |
| 09:19:09 | bauzas | maybe because the original commit was using a wrong section name ? | |
| 09:21:44 | gibi | yeah worth trying a new note file | |
| 09:21:51 | gibi | to see if that works or not | |
| 09:22:06 | bauzas | this is exactly what I'm testing :) | |
| 09:22:54 | bauzas | bingo | |
| 09:22:57 | opendevreview | Sylvain Bauza proposed openstack/nova master: DNM: Test issues https://review.opendev.org/c/openstack/nova/+/1007839 | |
| 09:23:42 | bauzas | we're stuck | |
| 09:24:05 | bauzas | basically because of the original section name, the relnote is not appearing | |
| 09:24:13 | bauzas | and I +2d that patch, f... | |
| 09:24:38 | bauzas | NEVER TRUST MY FYCKING BRAIN | |
| 09:30:48 | bauzas | hmmm, https://docs.openstack.org/reno/latest/releasenotes/index.html#relnotes-3-4-0 reno shouldn't have accepted the original YAML then | |
| 09:34:04 | bauzas | https://zuul.opendev.org/t/openstack/build/5648bd0f571e4304a63d2f39e1371283 used reno-4.1.0 so we should have failed it | |
| 09:37:40 | bauzas | OK, I think I understand the problem | |
| 09:38:31 | bauzas | we shipped the original file pre-rc1, which means we have the same commit on both master and 2026.2 branches but we're trying to update its master only | |
| 09:39:31 | bauzas | I honestly don't know how to fix that | |
| 09:40:59 | bauzas | stephenfin: tkajinam: you're unfortunately the last contributors of reno, so I probably need your help here :) | |
| 09:42:12 | bauzas | context : we had an original reno commit using a wrong section name that was merged before RC1, and now we want to fix it, but this isn't possible as the original commit is now branched (ie. https://review.opendev.org/c/openstack/nova/+/1007837/2 doesn't show up in the relnotes) | |
| 09:42:46 | bauzas | I guess we will need to backport that patch to 2026.2 which will automatically heal once we merge the two patches in tree, right? | |
| 09:43:04 | bauzas | I can try on my local git repo | |
| 09:49:49 | bauzas | gibi: Uggla: elodilles: OK, I get it, we need to merge some backport of https://review.opendev.org/c/openstack/nova/+/1007837 in 2026.2 before we tag 34.0.0 | |
| 09:50:28 | bauzas | I remembered reno using git tags | |
| 09:51:31 | opendevreview | Sylvain Bauza proposed openstack/nova stable/2026.2: Fix graceful shutdown release note section https://review.opendev.org/c/openstack/nova/+/1007840 | |
| 09:59:50 | opendevreview | Balazs Gibizer proposed openstack/nova master: Func test for cross cell resize with PCI source failure https://review.opendev.org/c/openstack/nova/+/1007841 | |
| 10:32:12 | opendevreview | Kamil Sambor proposed openstack/nova master: Replace StaticallyDelayingWrapper with futurist DelayingExecutor https://review.opendev.org/c/openstack/nova/+/997410 | |
| 11:25:58 | stephenfin | bauzas: iirc you shouldn't need to backport https://review.opendev.org/c/openstack/nova/+/1007837. reno will use the initial commit to determine what version to map to, but it should take the contents from the latest version of the file | |
| 11:34:28 | bauzas | stephenfin: but how to fix the issue ? | |
| 11:34:46 | bauzas | stephenfin: as you see the CI job doesn't create a right release note | |
| 11:58:36 | stephenfin | Sorry, lunch. I'd have expected https://review.opendev.org/c/openstack/nova/+/1007837 to fix it without the need for the backport but the CI result for that is pretty clear that it doesn't do so. I'm not sure why reno isn't picking that up 😕 | |
| 11:59:13 | bauzas | I'm pretty sure this is due to the original patch | |
| 11:59:27 | bauzas | it was tagged as a rc1 | |
| 12:00:05 | bauzas | I mean, a 2026.2 RC1 tag, so it will use the original section | |
| 12:00:16 | bauzas | so I guess we need to merge the backport before GA | |
| 12:01:57 | bauzas | stephenfin: gibi: Uggla: actually this works, yoohoo https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_d1d/openstack/d1dbcca71faf4c35a7d480cf66c2df82/docs/2026.2.html | |
| 12:02:12 | bauzas | I mean https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_d1d/openstack/d1dbcca71faf4c35a7d480cf66c2df82/docs/2026.2.html#known-issues | |
| 12:03:16 | bauzas | can you please +2/+W both of the patches ? | |
| 12:03:26 | bauzas | before GA | |
| 12:03:52 | bauzas | elodilles: are you OK with that ? | |
| 12:04:44 | stephenfin | I assume that CI result is from the backporT? | |
| 12:05:08 | stephenfin | btw, this isn't the only issue in our release notes. Run `.tox/releasenotes/bin/reno lint` locally... | |
| 12:06:49 | bauzas | stephenfin: sorry, indeed from https://zuul.opendev.org/t/openstack/build/d1dbcca71faf4c35a7d480cf66c2df82 | |
| 12:07:17 | bauzas | stephenfin: yeah I saw most of issues this morning b/c I ran reno lint | |
| 12:07:27 | elodilles | bauzas: with merging the reno fix to 2026.2 branch? | |
| 12:07:35 | bauzas | elodilles: yeah | |
| 12:07:44 | elodilles | yes, let's give it a try | |
| 12:08:09 | stephenfin | bauzas: I'm not sure reno can fix that because it would break everyone with currently broken releasenotes. Maybe with a new major version though? | |
| 12:08:17 | bauzas | elodilles: will we have a git tag for GA ? | |
| 12:08:53 | bauzas | stephenfin: sorry, about which issues ? | |
| 12:09:03 | elodilles | tomorrow, when we release, it will tag rc1 as final release, yes | |
| 12:09:12 | stephenfin | okay, got to the bottom of it. The backport is needed because the sphinx extension is only reading the specific branch | |
| 12:09:24 | bauzas | exactly | |
| 12:10:13 | bauzas | elodilles: stephenfin: actually, given reno will use rc1 as GA, I guess we don't need to merge the backport before GA, we can first merge the master change and then the stable one when we want | |
| 12:10:19 | elodilles | bauzas: though note, that it will tag rc1 as 34.0.0 | |
| 12:10:29 | elodilles | so not the tip of stable/2026.2 | |
| 12:10:53 | bauzas | elodilles: I guess we'll see the relnote in 34.0.0 even if we merge the backport after | |
| 12:10:53 | stephenfin | bauzas: there's zero risk to backporting this. I'd just merge it | |
| 12:11:08 | bauzas | as the original commit was merged before rc1 | |
| 12:11:08 | stephenfin | and yes, once the commit is on the branch it will appear. It doesn't need to be tagged | |
| 12:11:41 | stephenfin | arguably you don't even need the master commit. A stable-only one would likely do the job | |
| 12:11:42 | bauzas | but we definitely need to fix both the master and the stable/2026.2 changes | |
| 12:12:01 | bauzas | stephenfin: well, I'd prefer to fix both branches | |
| 12:12:20 | bauzas | if people look at existing relnotes, they would find some wrong section name | |
| 12:12:33 | stephenfin | I agree. I'm just noting it's not explicitly needed | |
| 12:12:43 | bauzas | yeah yeah I was knowing | |
| 12:12:46 | elodilles | bauzas: maybe we wait ~ an hour after the fix has landed on master branch, and we'll see | |
| 12:13:01 | bauzas | elodilles: cool | |
| 12:13:20 | bauzas | anyway, let's merge master first and we'll merge stable after | |
| 12:13:23 | elodilles | but anyway, I'm ok with merging the reno fix to stable/2026.2 | |
| 12:13:42 | bauzas | thanks stephenfin for +w'ing it btw. | |
| 12:14:09 | stephenfin | nw | |
| 12:14:16 | bauzas | elodilles: cool thanks, this will like any other backport | |
| 12:14:40 | elodilles | i've +2d it now | |
| 12:15:24 | bauzas | thanks | |
| 12:26:09 | gibi | OK I see both reno patch has the necessary votes | |
| 12:26:17 | gibi | thanks folks for handling that | |
| 13:05:50 | opendevreview | Merged openstack/nova master: Fix graceful shutdown release note section https://review.opendev.org/c/openstack/nova/+/1007837 | |
| 13:18:09 | elodilles | i've +W'd -> https://review.opendev.org/c/openstack/nova/+/1007840 | |
| 13:25:04 | gibi | elodilles: bauzas: thanks | |
| 13:27:21 | opendevreview | Sylvain Bauza proposed openstack/nova master: Prevent concurrent attach_volume with task_state guard https://review.opendev.org/c/openstack/nova/+/1006786 | |
| 13:27:21 | opendevreview | Sylvain Bauza proposed openstack/nova master: Guard against deleting a concurrent request's BDM on timeout https://review.opendev.org/c/openstack/nova/+/1005727 | |
| 13:27:22 | opendevreview | Sylvain Bauza proposed openstack/nova master: Reject stale reserve_block_device_name RPC on compute https://review.opendev.org/c/openstack/nova/+/1006011 | |
| 13:35:07 | elodilles | np | |