Github user srowen commented on the pull request:
https://github.com/apache/spark/pull/6586#issuecomment-108309863
Yes, but there are also advantages to letting it use byte buffers. They are
held in a thread-local soft reference so are cached and reused, and are
released under memory pressure. I don't think we want or need to manage this.
What's the problem, that this increases off-heap memory usage? yes, but you can
increase the amount of overhead YARN allows for this. Here you're just trading
for potentially less-efficient manual on-heap management. I am not clear this
is a good idea.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
