[core] Streaming generator task waits for all object report acks before finishing the task #44092
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
…re finishing the task (ray-project#44079) This fixes a bug in handling task (worker process) failures for streaming generator tasks. Streaming generator executors report the values of the dynamic returns to the caller asynchronously during the task execution. If a report is lost, then the caller will not be able to get the value of the yield'ed ObjectRef. Therefore, we need to wait until all in-flight object reports have been acked before we can finish the streaming generator task. Otherwise, it is possible for the caller to think that the task has finished, but then the executor dies before it can report the dynamic return object's value. Also cleans up some generator_waiter.cc headers, etc. Signed-off-by: Stephanie Wang <[email protected]>
stephanie-wang
requested review from
a team,
edoakes,
shrekris-anyscale,
zcin,
GeneDer,
akshay-anyscale,
matthewdeng,
justinvyu,
woshiyyya,
sven1977,
avnishn,
ArturNiederfahrenhorst,
maxpumperla,
kouroshHakha,
simonsays1980,
ericl,
scv119,
c21,
amogkam,
scottjlee,
bveeramani,
raulchen,
omatthew98,
kfstorm,
WangTaoTheTonic and
SongGuyang
as code owners
stephanie-wang
changed the title
|
@stephanie-wang we are pretty late into the release cycle. How critical is it tp idc this one into 2.10? |
We need it for Ray Data recovery. If we don't cherry-pick, I think we need to do a minor release at the very least. @c21 @raulchen can weigh in too. |
Yes, let's cherry pick the PR. |
|
yes, it's a critical stability fix |
zhe-thoughts
approved these changes
Mar 18, 2024
There was a problem hiding this comment.
@khluu Let's pick but be careful about possible issues (e.g. release test failures)
aslonnie
removed request for
a team,
edoakes,
zcin,
omatthew98,
bveeramani,
SongGuyang,
woshiyyya,
kouroshHakha,
avnishn,
shrekris-anyscale and
akshay-anyscale
can-anyscale
merged commit e032d7c
into
ray-project:releases/2.10.0
4 of 5 checks passed
|
merged thanks all |
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?
Cherry-pick #44079, fixes for #43914 and #43978.
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.