Page MenuHomePhabricator

1.38.0-wmf.2 deployment blockers
Closed, ResolvedPublicRelease

Details

Backup Train Conductor
dduvall
Release Version
1.38.0-wmf.2
Release Date
Sep 27 2021, 12:00 AM

2021 week 39 1.38-wmf.2 Changes wmf/1.38.0-wmf.2

This MediaWiki Train Deployment is scheduled for the week of Monday, September 27th:

Monday September 27thTuesday, September 28thWednesday, September 29thThursday, September 30thFriday
Backports only.Branch wmf.2 and deploy to Group 0 Wikis.Deploy wmf.2 to Group 1 Wikis.Deploy wmf.2 to all Wikis.No deployments on fridays

How this works

  • Any serious bugs affecting wmf.2 should be added as subtasks beneath this one.
  • Any open subtask(s) block the train from moving forward. This means no further deployments until the blockers are resolved.
  • If something is serious enough to warrant a rollback then you should bring it to the attention of deployers on the #wikimedia-operations IRC channel.
  • If you have a risky change in this week's train add a comment to this task using the Risky patch template
  • For more info about deployment blockers, see Holding the train.

Related Links

Other Deployments

Previous: 1.38.0-wmf.1
Next: 1.38.0-wmf.3

Event Timeline

thcipriani renamed this task from 1.37.0-wmf.25 deployment blockers to 1.38.0-wmf.2 deployment blockers.Aug 24 2021, 6:51 PM
thcipriani assigned this task to jeena.
thcipriani triaged this task as Medium priority.
thcipriani changed Release Version from 1.37.0-wmf.25 to 1.38.0-wmf.2.
thcipriani updated Other Assignee, added: dduvall.
Risky Patch! 🚂🔥
  • Change: https://gerrit.wikimedia.org/r/714883
  • Summary:
    • Timelines will now be rendered using BoxedCommand on Shellbox. This is risky because the Timeline extension is super sketchy.
  • Test plan:
    • Render example timelines on testwiki (I'll make sure I'm around during the train deploy)
  • Places to monitor:
  • Revert plan: Honestly, try to fix it live first, otherwise we can rollback Timeline to its 1.38.0-wmf.1 branch.
  • Affected wikis: all
  • IRC contact: legoktm
  • UBN Task Projects/tags: EasyTimeline

FYI: During the backport/config window, I just ran git fetch in the wmf.1 directory, which created all the wmf.2 remote-tracking branches there. Just so nobody’s confused when that wall of output doesn’t happen later. (The wmf.2 directory doesn’t exist yet and I’m not creating it – this is just about the remote tracking branches of the wmf.1 directory.)

Also, I just deployed this wmf.1 backport. I would also like to have it on wmf.2 (Gerrit change), but I’m not sure if merging into wmf.2 at the current point in time – after the branch was cut, but before it exists on deploy1002 – is okay, so I’m holding off for now. I’ll come back for it later, but if you want to, feel free to deploy it directly, it should be reasonably safe (and I already tested and deployed the wmf.1 version).

Change 724482 had a related patch set uploaded (by Jeena Huneidi; author: Jeena Huneidi):

[operations/mediawiki-config@master] testwikis wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724482

Change 724482 merged by jenkins-bot:

[operations/mediawiki-config@master] testwikis wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724482

Mentioned in SAL (#wikimedia-operations) [2021-09-28T17:55:41Z] <jhuneidi@deploy1002> Started scap: testwikis wikis to 1.38.0-wmf.2 refs T281166

@Jdlrobson closed subtask T291722: Text in Vector dropdown is too small (on Beta)

I suspect the inverse might happen in production, if not mitigated.

Ref https://gerrit.wikimedia.org/r/c/mediawiki/skins/Vector/+/723320, F34660692.

Should that mitigation block the train?

@Jdlrobson closed subtask T291722: Text in Vector dropdown is too small (on Beta)

I suspect the inverse might happen in production, if not mitigated.

Ref https://gerrit.wikimedia.org/r/c/mediawiki/skins/Vector/+/723320, F34660692.

Should that mitigation block the train?

This issue is limited to variants on wikis with variants as far as I can tell and on those the issue in question is the font size is a tiny bit larger than normal. A JavaScript patch could be provided to patch the font-size problem, but personally I don't see this as worth the effort involved given it doesn't meet the criteria of "Major stylistic problems affecting all pages " on https://wikitech-static.wikimedia.org/wiki/Deployments/Holding_the_train - product owner is @ovasileva and "ideally the product manager of the product with the regression should take responsibility for the decision"

Mentioned in SAL (#wikimedia-operations) [2021-09-28T18:45:07Z] <jhuneidi@deploy1002> Finished scap: testwikis wikis to 1.38.0-wmf.2 refs T281166 (duration: 49m 27s)

Change 724492 had a related patch set uploaded (by Jeena Huneidi; author: Jeena Huneidi):

[operations/mediawiki-config@master] group0 wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724492

Change 724492 merged by jenkins-bot:

[operations/mediawiki-config@master] group0 wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724492

Mentioned in SAL (#wikimedia-operations) [2021-09-28T19:08:21Z] <jhuneidi@deploy1002> rebuilt and synchronized wikiversions files: group0 wikis to 1.38.0-wmf.2 refs T281166

@Jdlrobson closed subtask T291722: Text in Vector dropdown is too small (on Beta)

I suspect the inverse might happen in production, if not mitigated.

Ref https://gerrit.wikimedia.org/r/c/mediawiki/skins/Vector/+/723320, F34660692.

Should that mitigation block the train?

This issue is limited to variants on wikis with variants as far as I can tell and on those the issue in question is the font size is a tiny bit larger than normal. A JavaScript patch could be provided to patch the font-size problem, but personally I don't see this as worth the effort involved given it doesn't meet the criteria of "Major stylistic problems affecting all pages " on https://wikitech-static.wikimedia.org/wiki/Deployments/Holding_the_train - product owner is @ovasileva and "ideally the product manager of the product with the regression should take responsibility for the decision"

+1 to this. If this is restricted to the size of language variants within the menu, it is not a significant enough regression to be a train blocker.

Change 724499 had a related patch set uploaded (by Jeena Huneidi; author: Jeena Huneidi):

[operations/mediawiki-config@master] group0 wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724499

Change 724499 merged by jenkins-bot:

[operations/mediawiki-config@master] group0 wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724499

Mentioned in SAL (#wikimedia-operations) [2021-09-28T19:48:12Z] <jhuneidi@deploy1002> rebuilt and synchronized wikiversions files: group0 wikis to 1.38.0-wmf.2 refs T281166

Someone who isn't me needs to figure out what to do with T280806: Remove old action api token parameters otherwise a bunch of bots are going to break when it rolls out to bigger wikis. I don't think current level of breakage (~66% of all Pywikibot-based bots) is acceptable, but it's really not my responsibility to keep reverting it.

@Legoktm just want to double check whether the above should be a blocker to rolling the train forward today?

Personally, I think so. But I've already reverted this twice and already said last week I wasn't going to again, someone else needs to take ownership of making that decision (which is why I didn't actually add it as a blocker).

Someone who isn't me needs to figure out what to do with T280806: Remove old action api token parameters otherwise a bunch of bots are going to break when it rolls out to bigger wikis. I don't think current level of breakage (~66% of all Pywikibot-based bots) is acceptable, but it's really not my responsibility to keep reverting it.

Personally, I think so. But I've already reverted this twice and already said last week I wasn't going to again, someone else needs to take ownership of making that decision (which is why I didn't actually add it as a blocker).

^ Asking @Pchelolo to take a look at this as the assignee of that task. I also think 66% of pywikibot based bots constitutes a major breakage, but I'm not in a position to know (a) if that's an explicit decision that anyone made (b) that will definitely happen.

Change 724833 had a related patch set uploaded (by Jeena Huneidi; author: Jeena Huneidi):

[operations/mediawiki-config@master] group1 wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724833

Change 724833 merged by jenkins-bot:

[operations/mediawiki-config@master] group1 wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/724833

Mentioned in SAL (#wikimedia-operations) [2021-09-29T20:21:32Z] <jhuneidi@deploy1002> rebuilt and synchronized wikiversions files: group1 wikis to 1.38.0-wmf.2 refs T281166

Mentioned in SAL (#wikimedia-operations) [2021-09-29T20:22:40Z] <jhuneidi@deploy1002> Synchronized php: group1 wikis to 1.38.0-wmf.2 refs T281166 (duration: 01m 08s)

I filed T292126 as a train blocker. If it should not block the train please correct me.

I filed T292126 as a train blocker. If it should not block the train please correct me.

It probably should just because it's breaking entire pages rather than just the <timeline> block, as it's supposed to. I have a fix ready https://gerrit.wikimedia.org/r/c/mediawiki/extensions/timeline/+/724855/ just trying to find someone who can CR +2 it.

Change 725093 had a related patch set uploaded (by Jeena Huneidi; author: Jeena Huneidi):

[operations/mediawiki-config@master] all wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/725093

Change 725093 merged by jenkins-bot:

[operations/mediawiki-config@master] all wikis to 1.38.0-wmf.2 refs T281166

https://gerrit.wikimedia.org/r/725093

Mentioned in SAL (#wikimedia-operations) [2021-09-30T19:14:06Z] <jhuneidi@deploy1002> rebuilt and synchronized wikiversions files: all wikis to 1.38.0-wmf.2 refs T281166

There is T292243: POI/marker disappeared on Wikivoyage maps generated on Toolforge, but I don't know if it is new in wmf.2.

Edit: This is by a toolforge tool and thus unrelated to the train.