zhoulii opened a new issue, #8408: URL: https://github.com/apache/paimon/issues/8408
### Search before asking - [x] I searched in the [issues](https://github.com/apache/paimon/issues) and found nothing similar. ### Paimon version 1.14 ### Compute Engine - ### Minimal reproduce step 1. Write a Parquet file with Paimon Java 2. Read the file with pyarrow 19.0.0 ### What doesn't meet your expectations? - Reading without filter fails with: OSError: Repetition level histogram size mismatch - Reading with filter may abort the Python process with a native core dump. ### Anything else? ## Root Cause This matches the following Apache Arrow issue: - [https://github.com/apache/arrow/issues/45283](https://github.com/apache/arrow/issues/45283) The issue title is: **[Python][C++][Parquet] "OSError: Repetition level histogram size mismatch" when reading parquet file in pyarrow since 19.0.0** In our native core dump, the failing stack also points to PyArrow C++ Parquet SizeStatistics validation: parquet::SizeStatistics::Validate(...) parquet::ColumnChunkMetaData::ColumnChunkMetaDataImpl(...) arrow::dataset::ParquetFileFragment::TestRowGroups(...) arrow::dataset::ParquetFileFragment::FilterRowGroups(...) So the failure happens when PyArrow 19.0.0 validates Parquet SizeStatistics metadata. The issue is fixed in the Arrow 19.0.1 milestone, and our local validation confirms that pyarrow==19.0.1 no longer reproduces the failure. ### Are you willing to submit a PR? - [x] I'm willing to submit a PR! -- 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]
