saths008 opened a new issue, #18554:
URL: https://github.com/apache/iceberg/issues/18554
### Feature Request / Improvement
This change avoids intermediate byte-array copies when writing
byte-array-backed Spark strings through Iceberg’s Parquet writer. It passes the
backing array, offset, and length directly to Parquet, while retaining the
copy-based fallback for other storage, including off-heap memory.
The bytes are marked as reused so Parquet copies values it retains,
preserving correctness when Spark reuses the backing memory.
Related: SPARK-60321 (https://issues.apache.org/jira/browse/SPARK-60321).
### Query engine
Spark
### Willingness to contribute
- [x] I can contribute this improvement/feature independently
- [ ] I would be willing to contribute this improvement/feature with
guidance from the Iceberg community
- [ ] I cannot contribute this improvement/feature at this time
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
