otterc commented on PR #36305: URL: https://github.com/apache/spark/pull/36305#issuecomment-1105642564
I don't think we can handle corruption when it happens in between a shuffle chunk and fallback because we don't know how much of the chunk has been consumed. We would have to treat this as a corruption of a regular shuffle block and propagate it to driver. So, on the driver now we will have to handle fetch failures related to a shuffle chunk. -- 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]
