Skip to content

Enable columnar shuffle by default #95

Description

@viirya

What is the problem the feature request solves?

COMET_COLUMNAR_SHUFFLE_ENABLED is the config Comet uses to decide if columnar shuffle should be used instead of native shuffle, if comet shuffle is enabled.

Because columnar shuffle covers more use cases (hash, range and single partitioning) than native shuffle. We should change this config to prefer columnar shuffle.

Note that this is blocked by the Java Arrow issue: apache/arrow#40038 (PR: apache/arrow#40043)

Currently if columnar shuffle is enabled, running Comet with TPCDS queries will get the error:

General execution error with reason org.apache.comet.CometNativeException: Fail to process Arrow array with reason C Data interface error: The external buffer at position 1 is null...

Describe the potential solution

No response

Additional context

No response

Activity

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

Metadata

Metadata

Assignees

Labels

enhancementNew feature or request

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions