Implement write.parquet.row-group-size-bytes in the pyarrow writer#12
Open
stephrb wants to merge 1 commit into
Open
Implement write.parquet.row-group-size-bytes in the pyarrow writer#12stephrb wants to merge 1 commit into
stephrb wants to merge 1 commit into
Conversation
The pyiceberg writer has historically ignored write.parquet.row-group-size-bytes (logging 'not implemented') and used only write.parquet.row-group-limit (rows). For wide tables that means a single row group ends up at gigabytes — e.g. 337 cols × 1,048,576 default rows ≈ 1.7 GiB uncompressed per row group — which drives the polars / pyarrow reader's decode peak into the tens of GiB on production reads. Now write_file resolves row_group_size as min(row_group_limit, row_group_size_bytes / bytes_per_row), where bytes_per_row is approximated from the in-memory arrow_table's nbytes. This matches Spark / parquet-mr 'whichever limit fires first' semantics and lets the existing PARQUET_ROW_GROUP_SIZE_BYTES_DEFAULT (128 MiB) actually take effect.
stephrb
force-pushed
the
sbuck/implement-row-group-size-bytes
branch
from
a707099 to
571647d
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The pyiceberg writer has historically ignored
write.parquet.row-group-size-bytes (logging 'not implemented') and used only write.parquet.row-group-limit (rows). For wide tables that means a single row group ends up at gigabytes — e.g. 337 cols × 1,048,576 default rows ≈ 1.7 GiB uncompressed per row group — which drives the polars / pyarrow reader's decode peak into the tens of GiB on production reads.
Now write_file resolves row_group_size as
min(row_group_limit, row_group_size_bytes / bytes_per_row), where bytes_per_row is approximated from the in-memory arrow_table's nbytes. This matches Spark / parquet-mr 'whichever limit fires first' semantics and lets the existing PARQUET_ROW_GROUP_SIZE_BYTES_DEFAULT (128 MiB) actually take effect.