Github user squito commented on the issue:
https://github.com/apache/spark/pull/21811
> I like it, but this will still create the byte channel right? is there a
way to reuse it?
we could create a pool, though management becomes a bit more complex.
would you ever shrink the pool, or would it always stay the same size? I guess
it shouldn't grow more than 1 byte array per io thread.
I'd rather get this simple fix in first before doing anything more
complicated ...
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
