Ergonomic, hermetic, user-friendly software build system for Python, Go, Java, Kotlin, Protobuf, Scala, Shell. Pants lets you fearlessly scale up your codebase!

Pants

I've got a bunch of `Filesystem has changed during run: retrying &lt;&gt; in 500ms...` messages in a loop while running `pants test ::`

The tasks that are being retried are "correct". I did change the filesystem in such a way that it would have caused those tasks to need to be redone. But it never stops. I even tried Ctrl+C out, wait a few minutes, and try again. And it just keeps looping a bunch of retry messages.

Did I get pants' filewatching into a glitched up state?
details in :thread:

I was running a large test suite, and after the first few failures, I could see what the issue was, so I fixed it and got stuck like this, where it keeps saying its detected the changes, but never actually gets to retrying anything.

Typically I would assume the watcher is too broad, and it's picking up temp files or cache changes (unless your tests actually write to disk)

Not sure how you got it into a state that it doesnt retry :exploding_head:

If you wipe the .pants.d folder and re-try, does it still happen?

I got it working again by `kill -9 &lt;pid of pantsd&gt;`

And it doesn't seem to be getting back into that state again. So :shrug: