[data] Trigger UDF.__del__ when an actor is about to exit
#49929
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: Hao Chen <[email protected]>
Signed-off-by: Hao Chen <[email protected]>
comaniac
approved these changes
Jan 18, 2025
srinathk10
approved these changes
Jan 22, 2025
bveeramani
approved these changes
Jan 22, 2025
Comment on lines
+413
to
+424
| def on_exit(self): | ||
| """Called when the actor is about to exist. | ||
| This enables performing cleanup operations via `UDF.__del__`. | ||
|
|
||
| Note, this only ensures cleanup is performed when the job exists gracefully. | ||
| If the driver or the actor is forcefully killed, `__del__` will not be called. | ||
| """ | ||
| # `_map_actor_context` is a global variable that references the UDF object. | ||
| # Delete it to trigger `UDF.__del__`. | ||
| del ray.data._map_actor_context | ||
| ray.data._map_actor_context = None | ||
|
|
There was a problem hiding this comment.
Is there a reason we need to introduce an explicit method rather than overriding _MapWorker.__del__? I imagine if we override __del__, we'll implicitly do the cleanup when we run del self._running_actors[actor].
There was a problem hiding this comment.
I also tried that. But it looks like that Ray Core doesn't call Actor.__del__ when killing an actor. cc @jjyao
Signed-off-by: Hao Chen <[email protected]>
erictang000
pushed a commit
to erictang000/ray
that referenced
this pull request
Jan 27, 2025
…ect#49929) ## Why are these changes needed? A stateful UDF may depend on external resources that may need to be cleaned up at the end. This PR allows performing the cleanup operations via `UDF.__del__` Note, this only ensures cleanup is performed when the job exists gracefully. If the driver or the actor is forcefully killed, `__del__` will not be --------- Signed-off-by: Hao Chen <[email protected]> Signed-off-by: Eric Tang <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Why are these changes needed?
A stateful UDF may depend on external resources that may need to be cleaned up at the end.
This PR allows performing the cleanup operations via
UDF.__del__Note, this only ensures cleanup is performed when the job exists gracefully.
If the driver or the actor is forcefully killed,
__del__will not be called.Related issue number
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.method in Tune, I've added it in
doc/source/tune/api/under thecorresponding
.rstfile.