Spark: Use actual file sizes instead of schema-based estimates for table statistics#15693
Open
majian1998 wants to merge 2 commits intoapache:mainfrom
Open
Spark: Use actual file sizes instead of schema-based estimates for table statistics#15693majian1998 wants to merge 2 commits intoapache:mainfrom
majian1998 wants to merge 2 commits intoapache:mainfrom
Conversation
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
SparkScan.estimateStatistics() currently estimates table size by multiplying StructType.defaultSize() (hardcoded per-type constants, e.g. STRING=54 bytes) by the total row count. This can be wildly inaccurate compared to actual data, causing Spark to pick suboptimal join strategies (e.g. missing BroadcastHashJoin or broadcasting a table that's too large).
Changes
Replace the type-based estimation with real file size data that Iceberg already tracks:
Partitioned tables (no filters): read total-files-size from snapshot summary
All other paths: sum ScanTaskGroup.sizeBytes() which reflects actual fileSizeInBytes from manifests
Applies to both SparkScan and SparkChangelogScan
This makes Iceberg-based table statistics consistent with how Spark's native Parquet source reports size (using actual file sizes on disk), so the same data produces the same join strategy regardless of the source.
Related issue: #15684