| *** tosky has quit IRC | 00:10 | |
| *** icey has quit IRC | 00:23 | |
| masayukig | hi release team, we're struggling with reno. It seems that reno can't find the 25.0.0 tag in tempest. Any thoughts on this? https://review.opendev.org/c/openstack/tempest/+/752786 | 00:27 |
|---|---|---|
| openstackgerrit | Steve Baker proposed openstack/releases master: Release metalsmith 1.4.0 for wallaby https://review.opendev.org/c/openstack/releases/+/767060 | 00:48 |
| *** armax has quit IRC | 02:39 | |
| *** ykarel has joined #openstack-release | 04:26 | |
| *** vishalmanchanda has joined #openstack-release | 05:04 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-release | 05:33 | |
| *** brinzhang has joined #openstack-release | 05:55 | |
| *** brinzhang has quit IRC | 07:10 | |
| *** brinzhang has joined #openstack-release | 07:10 | |
| *** icey has joined #openstack-release | 07:42 | |
| *** sboyron has joined #openstack-release | 07:55 | |
| *** slaweq has quit IRC | 07:56 | |
| *** slaweq has joined #openstack-release | 07:58 | |
| *** rpittau|afk is now known as rpittau | 08:06 | |
| hberaud | masayukig: I looking for details | 08:26 |
| masayukig | hberaud: thank you so much! | 08:27 |
| hberaud | masayukig: you're welcome | 08:28 |
| hberaud | masayukig: apparently you haven't even reno for victoria https://docs.openstack.org/releasenotes/tempest/victoria.html | 08:30 |
| hberaud | masayukig: ah wait... it's maybe because you are branchless | 08:32 |
| masayukig | hberaud: yeah, tempest is branchless | 08:34 |
| hberaud | masayukig: the logs of build-openstack-releasenotes ( https://zuul.opendev.org/t/openstack/build/314be05d54ef429d85aa5d4fdbcbec9d ) are too old, I can't access them | 08:35 |
| hberaud | masayukig: can we retry to rebase this patch (by example) to trigger a new execution and get fresh logs? | 08:36 |
| masayukig | hberaud: sorry about that. yeah, it should be OK | 08:37 |
| hberaud | masayukig: no problem | 08:37 |
| masayukig | hberaud: just rebased | 08:40 |
| hberaud | masayukig: thanks | 08:41 |
| *** tosky has joined #openstack-release | 08:41 | |
| hberaud | masayukig: let's wait a bit for fresh logs | 08:42 |
| masayukig | hberaud: yeah, btw, FWIW, I dug a bit on the comment. But I'm still 100% not sure. https://review.opendev.org/c/openstack/tempest/+/752786/2#message-150bc4adf0ca3d82c2068eaad87f7c0045306dea | 08:43 |
| * hberaud read | 08:45 | |
| hberaud | masayukig: apparently you're right only merge commits are retrieved | 08:49 |
| hberaud | masayukig: all these SHAs seems to matches with the outcome of "git log --oneline --merges " | 08:51 |
| masayukig | hberaud: yeah, is that expected behavior? | 08:51 |
| hberaud | masayukig: I don't know sorry... maybe someone more involved than me within reno (dhellmann by example) can answer your question | 08:54 |
| masayukig | hberaud: thanks, no worries. Do you think it's good to bring this to the openstack-discuss ML? | 08:56 |
| hberaud | masayukig: hm maybe we can wait for reply of someone more experienced on reno before triggering a ML discussion | 08:58 |
| hberaud | masayukig: also I'll try to dig a bit into reno before | 08:58 |
| masayukig | hberaud: thanks! sure. I'm going to call it a day soon, actually. But I'll check the irc log later | 09:01 |
| hberaud | ack thanks | 09:07 |
| hberaud | masayukig: here is the delta of changes added to reno since your latest release note on tempest => http://paste.openstack.org/show/801041/ | 09:15 |
| *** e0ne has joined #openstack-release | 09:15 | |
| *** vishalmanchanda has quit IRC | 10:22 | |
| *** rpittau is now known as rpittau|bbl | 10:27 | |
| *** vishalmanchanda has joined #openstack-release | 10:30 | |
| *** dtantsur|afk is now known as dtantsur | 11:00 | |
| openstackgerrit | Riccardo Pittau proposed openstack/releases master: Release bifrost 10.0.0 for wallaby https://review.opendev.org/c/openstack/releases/+/767121 | 12:00 |
| *** ricolin has quit IRC | 12:09 | |
| *** ykarel_ has joined #openstack-release | 12:14 | |
| *** ykarel has quit IRC | 12:17 | |
| openstackgerrit | Hervé Beraud proposed openstack/releases master: Replace series name by template variable https://review.opendev.org/c/openstack/releases/+/767128 | 12:44 |
| openstackgerrit | Hervé Beraud proposed openstack/releases master: Remove useless canary tests https://review.opendev.org/c/openstack/releases/+/767129 | 12:52 |
| *** rpittau|bbl is now known as rpittau | 13:12 | |
| *** brinzhang has quit IRC | 13:21 | |
| *** brinzhang has joined #openstack-release | 13:22 | |
| *** brinzhang has quit IRC | 13:23 | |
| *** brinzhang has joined #openstack-release | 13:23 | |
| *** ykarel_ is now known as ykarel | 13:23 | |
| dhellmann | masayukig , hberaud : I think I missed part of the conversation about reno. What is happening? | 13:48 |
| *** vishalmanchanda has quit IRC | 13:52 | |
| hberaud | dhellmann: apparently the releasenotes job fail on tempest with the latest version that they try to generate https://review.opendev.org/c/openstack/tempest/+/752786 | 13:54 |
| hberaud | dhellmann: they got timeout due to something like => 'unable to find release notes file associated with unique id '27000c02edc5a112', skipping' https://zuul.opendev.org/t/openstack/build/3834bae67c3f43b0af82ae525c6eb652/log/job-output.txt | 13:55 |
| dhellmann | is there a tag with that version number? | 13:56 |
| hberaud | dhellmann: yes the tag exist | 13:57 |
| dhellmann | this error means that sphinx didn't find it on the branch it was told to scan https://zuul.opendev.org/t/openstack/build/3834bae67c3f43b0af82ae525c6eb652/log/job-output.txt#648 | 13:57 |
| hberaud | dhellmann: https://opendev.org/openstack/tempest/src/tag/25.0.0 | 13:57 |
| dhellmann | oh, no, that's not right. it did find it, there just aren't any release notes | 13:57 |
| *** jbadiapa has joined #openstack-release | 13:58 | |
| dhellmann | let me see if I can reproduce this | 13:58 |
| hberaud | dhellmann: ack. FYI this morning I asked them to rebase that patch to trigger a new execution of that job and get fresh logs and [1] is the outcome of that [1] https://zuul.opendev.org/t/openstack/build/3834bae67c3f43b0af82ae525c6eb652/log/job-output.txt | 14:00 |
| hberaud | dhellmann: previous logs expired and weren't available anymore | 14:00 |
| hberaud | dhellmann: thanks for your help | 14:01 |
| openstackgerrit | Ghanshyam proposed openstack/releases master: Release Tempest 26.0.0 https://review.opendev.org/c/openstack/releases/+/767163 | 14:32 |
| openstackgerrit | Ghanshyam proposed openstack/releases master: Release Patrole 0.11.0 https://review.opendev.org/c/openstack/releases/+/767164 | 14:34 |
| *** sboyron has quit IRC | 14:56 | |
| *** ykarel has quit IRC | 15:01 | |
| dhellmann | hberaud, masayukig : the 25.0.0 tag is not on the master branch, it is on a commit that was merged into the master branch. reno therefore does not consider it as "inline" between 24.0.0 and HEAD so it does not put any of the release notes into that version | 15:04 |
| dhellmann | I thought the release machinery prevented this sort of error but I guess not | 15:04 |
| hberaud | dhellmann: good to know that | 15:04 |
| dhellmann | in any case, I suggest tagging 25.0.1 on 94a76091f | 15:05 |
| *** armax has joined #openstack-release | 15:06 | |
| hberaud | ack, thanks for these details, FYI gmann ^ | 15:06 |
| gmann | dhellmann: hberaud i see thanks. I will push 25.0.1 on 94a76091f | 15:12 |
| hberaud | gmann: ack, thanks | 15:12 |
| *** sboyron has joined #openstack-release | 15:14 | |
| gmann | dhellmann: hberaud but i did not get 'the 25.0.0 tag is not on the master branch, it is on a commit that was merged' | 15:15 |
| gmann | it is released on c8d7a01014d4b82dd9960679db5e77dc31f4d87a right | 15:16 |
| hberaud | gmann: yes and master contains c8d7a01014d4b82dd9960679db5e77dc31f4d87a | 15:16 |
| dhellmann | if you look at the 'git log --decorate --graph' output and search for the tag, you'll see that it's on a commit that is then merged into the master branch. reno's toplogical sort isn't smart enough to realize that that means it is on the branch it was merged into | 15:16 |
| dhellmann | the other commit I mention is the merge commit where the tagged commit is brought into the master branch | 15:17 |
| hberaud | if I correctly understand the release should have been on 94a76091f616a4e4a46534b82d4974534d197643 | 15:18 |
| hberaud | to avoid this situation, right? | 15:19 |
| hberaud | where 94a76091f616a4e4a46534b82d4974534d197643 is the merge commit related to c8d7a01014d4b82dd9960679db5e77dc31f4d87a | 15:19 |
| dhellmann | yes that's right | 15:20 |
| dhellmann | it is OK to tag the left-hand child of a merge, but not the right-hand child | 15:20 |
| hberaud | ok make sense to me | 15:20 |
| hberaud | I'll take a further look to our mechanismes to see why we didn't see that | 15:21 |
| gmann | dhellmann: hberaud but does it cause issue always? I remember we might have release few old version with 'commit on master' ? | 15:24 |
| dhellmann | if any of the other versions were tagged this way they would have produced the same error | 15:24 |
| hberaud | however I don't expect that we can rely on the related jobs logs (https://review.opendev.org/c/openstack/releases/+/752665) as this was released 2 months ago and the logs expired too | 15:25 |
| dhellmann | this behavior is a fundamental part of some of the oldest code in reno, so it's not a new change. the reason you're seeing it now is that you're explicitly asking for details about 25.0.0 with the changes in https://review.opendev.org/c/openstack/tempest/+/752786 | 15:25 |
| dhellmann | if you run `reno report .` from a clean checkout of tempest it does not show the 25.0.0 version even though the tag is there, because reno does not consider any of the release notes to be "in" that release | 15:26 |
| dhellmann | that KeyError should be wrapped with a more clear message about having been asked for a version that has no data | 15:27 |
| hberaud | on the release side I'll try to reproduce a similar scenario locally or a "DNM" patch to see what's happen and see if we can raise some warnings in this kind of situation | 15:28 |
| gmann | Tempest 22.0.0 is released same way on 13c190d62e4c4596339eeee93eb0425cd525d71f but reno did not cause issue | 15:28 |
| dhellmann | I really thought the validation code had logic to prevent this because we had some early releases early on that came from the wrong branch. It may only be looking at the output of `git branch --contains` though, and that's not sufficient for this specific error | 15:29 |
| hberaud | ack | 15:29 |
| dhellmann | gmann : in the graph output 22.0.0 is not the right-hand child of a merge | 15:30 |
| dhellmann | see the difference in the graph structure in http://paste.openstack.org/show/801062/ | 15:31 |
| dhellmann | line 10 is the 22.0.0 tag and line 22 is 25.0.0 | 15:32 |
| dhellmann | you only ever want to tag releases from commits with the * in the left column of the graph output | 15:32 |
| gmann | dhellmann: and that depends on the timing of the release (if we release with the 'commit on master')? that is why 22.0.0 is on left-hand child ? | 15:35 |
| dhellmann | if you always tag HEAD, it will be fine. if you tag an "old" commit, you have to be careful | 15:36 |
| gmann | i see. | 15:36 |
| dhellmann | that's why the new-release command in the release repo defaults to whatever it sees as HEAD at the point that you run it | 15:37 |
| *** markmcclain has joined #openstack-release | 15:38 | |
| *** jbadiapa has quit IRC | 15:40 | |
| *** jbadiapa has joined #openstack-release | 15:41 | |
| gmann | dhellmann: thanks for the clarification. I think this is one good thing to learn today :) | 15:44 |
| dhellmann | congratulations on finding an obscure bug! ;-) | 15:45 |
| gmann | :) | 15:45 |
| gmann | just wondering if we can autodetect this in in release tooling ? | 15:45 |
| dhellmann | I expect so. Like I said above, I think the check today is using `git branch --contains` to ensure that a commit being tagged is on the right branch, but we would need another check to verify the relative position of that commit and any merge commit that pulls it in, if there is one | 15:47 |
| *** sboyron has quit IRC | 15:51 | |
| *** sboyron has joined #openstack-release | 15:52 | |
| openstackgerrit | Ghanshyam proposed openstack/releases master: Release Tempest 25.0.1 https://review.opendev.org/c/openstack/releases/+/767180 | 15:53 |
| *** e0ne has quit IRC | 17:11 | |
| *** dtantsur is now known as dtantsur|afk | 17:29 | |
| *** rpittau is now known as rpittau|afk | 17:46 | |
| *** e0ne has joined #openstack-release | 18:48 | |
| *** e0ne has quit IRC | 20:22 | |
| *** sboyron has quit IRC | 21:02 | |
| *** e0ne has joined #openstack-release | 21:02 | |
| *** e0ne has quit IRC | 21:03 | |
| *** jbadiapa has quit IRC | 21:06 | |
| *** trident has quit IRC | 21:56 | |
| *** trident has joined #openstack-release | 22:00 | |
| masayukig | dhellmann: hberaud gmann thanks for the clarification! | 23:00 |
| masayukig | gmann: thanks. +1 for the 25.0.1 patch | 23:05 |
| *** slaweq has quit IRC | 23:32 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!