[core] Always make request on fetch_local in wait even if num_objects in memory #50121
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: dayshah <[email protected]>
dayshah
force-pushed
the
always-wait-fetch-local
branch
from
ebc96c1 to
04eb297
Compare
Signed-off-by: dayshah <[email protected]>
dayshah
commented
Jan 31, 2025
| @@ -14,6 +14,7 @@ COPTS_WITHOUT_LOG = select({ | |||
| "//conditions:default": [ | |||
| "-Wunused-result", | |||
| "-Wconversion-null", | |||
| "-Wmisleading-indentation", | |||
There was a problem hiding this comment.
jjyao
reviewed
Jan 31, 2025
Signed-off-by: dayshah <[email protected]>
dayshah
commented
Feb 3, 2025
| # to start pulling y and z | ||
| ray.wait([x, y, z], num_returns=1, fetch_local=True) | ||
| time.sleep(3) | ||
|
|
There was a problem hiding this comment.
memory_summary and list_objects show that objects are all always in the same store? Is making fake nodes using cluster.add_node not a viable solution for this, do we need a test env with two actual nodes
Signed-off-by: dayshah <[email protected]>
jjyao
reviewed
Feb 3, 2025
Comment on lines
+107
to
+111
| start_time = time.perf_counter() | ||
| ray.get([y, z]) | ||
| # y and z should be immediately available as pull requests should've | ||
| # been made immediately on the ray.wait call | ||
| time_to_get = time.perf_counter() - start_time |
There was a problem hiding this comment.
You can check ray.available_resources()
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Outdated suggestions cannot be applied.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?
We always want to make pull requests to get the objects passed into
ray.waiton different nodes whenray.waitis called withfetch_local. Right now we don't make that request if we've already gottennum_objectsfrom the core worker memory store.Related issue number
Closes #49257
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.