[C++][Parquet] Make compression adaptive with V2 data pages
Open
Component: C++
Component: Parquet
good-second-issue
Type: enhancement
- Dominant language
- C++
- Stars
- 17.1k
- Forks
- 4.3k
- Avg merge
- 3d 13h
- Merged PRs (30d)
- 88
Description
### Describe the enhancement requested
In [BuildDataPageV2](https://github.com/apache/arrow/blob/33d1f32136f950520dcca278b77acff430dc8dca/cpp/src/parquet/column_writer.cc#L1045-L1053), we enable compression as soon as the compressed size is smaller than the uncompressed size.
However, this adds overhead when reading even when the compression gain is very small. We could probably be slightly smarter and make the threshold configurable.
### Component(s)
C++, Parquet
Contributor guide
Assessment
This issue has not been assessed yet.