fix: Avoid creating huge duplicate of canonicalized plans for CometNativeExec #639
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
This reverts commit b272551.
viirya
changed the title
andygrove
approved these changes
Jul 8, 2024
|
Merged. Thanks @andygrove |
himadripal
pushed a commit
to himadripal/datafusion-comet
that referenced
this pull request
Sep 7, 2024
…tiveExec (apache#639) * fix: Remove original plan parameter from CometNativeExec * Revert "fix: Remove original plan parameter from CometNativeExec" This reverts commit b272551. * More * Revert "Revert "fix: Remove original plan parameter from CometNativeExec"" This reverts commit 722dc07. * More * More * Fix * Fix diffs * Update
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Outdated suggestions cannot be applied.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.
Which issue does this PR close?
Closes #594.
Rationale for this change
I encountered OOM issue when trying to enable columnar shuffle for TPCDS queries in #613. I ran memory profiling and found that a lot of Spark expressions are allocated, e.g.,
AttributeReference:It seems caused by the recursively canonicalization in
canonicalizePlanson the original plan parameter. I think it is because when we transform the query plan, original plans in each operator could reference mix of Spark and Comet plans. When the canonical plans are required, it produces huge duplicate of canonicalized plans for each query snippets.We don't need to rely on the original plan to compare canonical plans of Comet. This patch simply changes the canonical plan of the original plan to null.
What changes are included in this PR?
How are these changes tested?