wForget commented on issue #9959:
URL: 
https://github.com/apache/incubator-gluten/issues/9959#issuecomment-3017791612

   > [@wForget](https://github.com/wForget) without 
[#9278](https://github.com/apache/incubator-gluten/pull/9278), we need to 
compress the whole spilled data once, which needs to allocate a large offheap 
memory. With [#9278](https://github.com/apache/incubator-gluten/pull/9278), we 
introduced streaming compression which has a fixed buffer size allocating in 
overhead memory instead of offheap memory. In this way we will compress the 
spilled data into several small blocks. This is the way vanilla spark uses.
   > 
   > Can you try and report any issues?
   
   Thanks, I will try to implement it in hash-based shuffle.
   


-- 
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]

Reply via email to