You are viewing a plain text version of this content. The canonical link for it is here.

Posted to github@arrow.apache.org by "justinli500 (via GitHub)" <gi...@apache.org> on 2026/04/24 05:53:33 UTC

Re: [I] [Python] Dataset.to_batches accumulates memory usage and leaks [arrow]

justinli500 commented on issue #39808:
URL: https://github.com/apache/arrow/issues/39808#issuecomment-4311007928

   I've been working on a fix at the C++ Parquet reader level - evicting pre-buffered row-group bytes after they're consumed, so pre_buffer=True doesn't retain the full file in memory during iteration. It's a different approach from the recent scanner/Ray-side work (ray-project/ray#62745, b234d4e).
   
   Opening a PR shortly - happy to coordinate if there's overlap I'm missing.


-- 
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: github-unsubscribe@arrow.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org