steveloughran commented on pull request #2179:
URL: https://github.com/apache/hadoop/pull/2179#issuecomment-686411091
LGTM, +1 *pending you make clear in the option names and docs that this is
experimental*
> These configs are tested on prod environments.
> The same can
steveloughran commented on pull request #2179:
URL: https://github.com/apache/hadoop/pull/2179#issuecomment-680982644
Here is my current view of this patch
* as a quick workaround for scale-up issues in the existing code -it works
* as the long term design for something which
steveloughran commented on pull request #2179:
URL: https://github.com/apache/hadoop/pull/2179#issuecomment-680270262
> We are working on a similar approach. The same needs extensive tests to
ensure the perf side is fine. Till the fix is available this can be used to
control the memory
steveloughran commented on pull request #2179:
URL: https://github.com/apache/hadoop/pull/2179#issuecomment-678634936
Prefer you use the strategy outlined in HADOOP-17195. Rather than pool
settings for each output stream, the store has a single shared pool with
semaphores to limit #of