-
-
Notifications
You must be signed in to change notification settings - Fork 719
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Partial revert of compute-task message format #6626
Partial revert of compute-task message format #6626
Conversation
@jakirkham maybe you have an idea what's going on here? |
with pytest.raises(IndexError): | ||
overlapped.compute() |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is obviously not ideal but I simply took the reproducer from #6624
I do not hit the serialization error anymore but rather an IndexError ¯_(ツ)_/¯
Unit Test ResultsSee test report for an extended history of previous test failures. This is useful for diagnosing flaky tests. 15 files ±0 15 suites ±0 10h 8m 29s ⏱️ - 24m 59s Results for commit 27f9f67. ± Comparison against base commit dc019ed. ♻️ This comment has been updated with latest results. |
All green 🎉 👀 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This seems like a nice clean solution to me!
"run_spec": None, | ||
"function": None, | ||
"args": None, | ||
"kwargs": None, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't 100% see the point of initializing these all with None but I also don't see the harm in it.
Shall I go ahead and merge this @fjetter or do you think it needs another set of eyes? |
merging |
This is a hotfix for #6624 by reverting the compute-task message format almost to the original state before #6410
I didn't debug very deeply but it appears that our (de-)serializer cannot handle nested Serialized objects very well or something else is messed up. By formatting the
run_spec
value as a dict the way #6410 did appears to be a compatible change but the nesting is destroyed as soon as it reaches the worker.basically