westonpace commented on PR #41335:
URL: https://github.com/apache/arrow/pull/41335#issuecomment-2089189076

   > Yes, this is my assumption. And the overhead is actually 64MB per plan * 
per thread.
   
   Ah, good point, 64MB per thread per plan is too much.
   
   > Per-node (meanwhile per-thread, presumably) temp stack fixes this type of 
overflow by effectively being able to implicitly expand the overall temp stack 
size by a multiple of the number of nodes.
   
   I agree that per-node stacks are a reasonable solution.
   I think Antoine's solution is also a reasonable solution.
   
   I'd be happy with either approach.


-- 
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]

Reply via email to