MODULE.bazel also needs updating. That is why the Bazel test is failing.

Done.

justFYI: 4.1.111.Final would be out https://netty.io/news/2024/06/11/4-1-111-Final.html

justFYI: 4.1.111.Final would be out https://netty.io/news/2024/06/11/4-1-111-Final.html

I reported #11284 about 4.1.111.Final. It would be useful to run CI with it to see what breaks.

@ejona86 Windows test is timing out. Any idea about this test?

It can flake. Follow the link to Test Fusion and click the "rebuild" button.

This needs to be backported to the 1.66 branch. Easiest way is to use https://gist.github.com/sergiitk/af606cbab72b00d8db8a53d21e766ee9 (./backport.sh grpc-java 11273 66). It does some niceties like adding "(v1.66.x backport)" to the PR description, adding a reference back to the original PR so you can see the list of all backports in one location. But it is also essentially just a cherry-pick onto a branch, so isn't mandatory.

Reverting in #11440

Once I noticed there were unintended commits that caused test failure I panic reverted them but I didn't notice there were more commits to revert (there were 4 but I only noticed 2) so my revert left things in a half baked state.

I had thought once certain changes are committed to a branch https://screenshot.googleplex.com/86pkpKBjbbagzvR they belong to that branch. But what is happening is that they get carried over when I switch branch and get committed again with a different hash. I don't really understand how this works. I know commits carry over if I do checkout -b but I only checked out an existing branch in this case. @ejona86

Commits stay on branches, but certain operations can copy commits between branches, like rebase and merge. I could probably look at your bash history to figure out how it happened. It does seem like it is happening very frequently for you.