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.

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

[CORE][SPARK-4.0] Only require one config spark.plugins to enable gluten #5915

Open
ulysses-you opened this issue May 29, 2024 · 4 comments
Open
Labels
enhancement New feature or request

Comments

@ulysses-you
Copy link
Contributor

Description

For now, there are at least three required configs to enable gluten:

  • spark.plugins
  • spark.memory.offHeap.enabled
  • spark.shuffle.manager

Less required configs is always better. Fortunately, We can set spark.memory.offHeap.enabled and spark.shuffle.manager inside GlutenDriverPlugin since Spark 4.0.0, so only spark.plugins is required.

@ulysses-you ulysses-you added the enhancement New feature or request label May 29, 2024
@ulysses-you ulysses-you changed the title Only require one config spark.plugins to enable gluten [CORE][SPARK-4.0] Only require one config spark.plugins to enable gluten May 29, 2024
@FelixYBW
Copy link
Contributor

We may set shuffle manager default as local shuffle manager. We may overwrite offheap.enabled=true once offheap.size is set.

@acvictor
Copy link
Contributor

acvictor commented Jun 3, 2024

@FelixYBW as long as ColumnarShuffleManager is in the classpath, do you see any issues with setting default shuffle manager to ColumnarShuffleManager?

@FelixYBW
Copy link
Contributor

FelixYBW commented Jun 3, 2024

@FelixYBW as long as ColumnarShuffleManager is in the classpath, do you see any issues with setting default shuffle manager to ColumnarShuffleManager?

No. go ahead to submit a PR.

@acvictor
Copy link
Contributor

acvictor commented Jun 4, 2024

@FelixYBW as long as ColumnarShuffleManager is in the classpath, do you see any issues with setting default shuffle manager to ColumnarShuffleManager?

No. go ahead to submit a PR.

Where would this be set? I think in Spark 3.x it must be set as part of session builder?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

3 participants