Re: [I] [SPARK-38101] execuors fail fetching map statuses with `INTERNAL_ERROR_BROADCAST` [spark]

["EnricoMi (via GitHub)" <gi...@apache.org>]

EnricoMi commented on issue #54723:
URL: https://github.com/apache/spark/issues/54723#issuecomment-4251779926

   I have created a pull request that makes Spark retry the task (which will fetch latest map statuses) rather than retrying the whole stage (or have the task retry fetching map statuses). This is also able to handle another source of internal broadcast error occurring when task binary broadcast variable cannot be fetched. That code path cannot be retried by the executor.


-- 
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: reviews-unsubscribe@spark.apache.org

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


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org