Skip to content
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.

Sign up for GitHub

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

Jump to bottom

PERF: get all partition widths/lengths in parallel instead of serially. #4494

Closed
mvashishtha opened this issue May 25, 2022 · 3 comments · May be fixed by #4683
Closed

PERF: get all partition widths/lengths in parallel instead of serially. #4494

mvashishtha opened this issue May 25, 2022 · 3 comments · May be fixed by #4683
Assignees
@noloerino
Labels
Performance 🚀 Performance related issues and pull requests.

Comments

@mvashishtha
Copy link
Collaborator

The reproducing script is the same as in #4493, but the solution here is different: instead of getting all lengths/widths serially, we should do so in parallel. We will need to add a new method at the physical layer for that.

This solution will not save us the cost of serializing the call queue drain result after computing length, but it will let us get all the lengths/widths at once instead of serially.
@mvashishtha mvashishtha mentioned this issue May 25, 2022
Closed
@mvashishtha mvashishtha added the Performance 🚀 Performance related issues and pull requests. label May 25, 2022
@mvashishtha mvashishtha mentioned this issue Jun 21, 2022
Merged
8 tasks
@noloerino noloerino self-assigned this Jul 11, 2022
@noloerino
Copy link
Collaborator

I see this is already implemented for Dask by #4420. Is adding a similar implementation for PandasOnRay sufficient, or do we also need similar changes for PyarrowOnRay and OmnisciOnNative as well?

@mvashishtha
Copy link
Collaborator Author

@noloerino here are all the places I can find where we are getting uncached partition shapes serially. I found them by searching for width() in the Modin codebase.

We do get lengths serially in rebalance_partitions, but that's okay because we check first that the lengths are cached.

@noloerino noloerino mentioned this issue Jul 18, 2022
Draft
8 tasks
noloerino added a commit to noloerino/modin that referenced this issue Aug 9, 2022
@noloerino
PERF-modin-project#4494: Get all partition widths/lengths in parallel
cb4f35c 
Signed-off-by: Jonathan Shi <[email protected]>
@mvashishtha
Copy link
Collaborator Author

It's hard to find a case where this optimization is useful. See my comment here: #4683 (comment)

Given that this optimization doesn't seem to give any major gains, it doesn't seem to be worth the extra code complexity in #4683. I'll close this issue for now.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@noloerino noloerino

Labels
Performance 🚀 Performance related issues and pull requests.
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

PERF-#4494: Get partition widths/lengths in parallel instead of serially noloerino/modin
2 participants
@mvashishtha @noloerino