Github user mateiz commented on the pull request:
https://github.com/apache/spark/pull/1165#issuecomment-49831796
BTW for these changes it would be good to perf-test it, e.g. design a
workload where your blocks do fit in memory (just on one multicore machine is
fine) and make sure there's no slowdown. But I think that if we do the doubling
we're in good shape. It's just calling synchronized every 1000 elements that
would be problematic.
---
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.
---
