okumin commented on code in PR #4432:
URL: https://github.com/apache/hive/pull/4432#discussion_r1236895217
##########
ql/src/java/org/apache/hadoop/hive/ql/parse/GenTezUtils.java:
##########
@@ -130,8 +130,8 @@ public static ReduceWork createReduceWork(
maxPartition = (maxPartition > maxReducers) ? maxReducers : maxPartition;
// reduce only if the parameters are significant
- if (minPartition < maxPartition &&
- nReducers * minPartitionFactor >= 1.0) {
+ final int parallelismThreshold =
context.conf.getIntVar(HiveConf.ConfVars.TEZ_AUTO_REDUCER_PARALLELISM_THRESHOLD);
+ if (minPartition < maxPartition && maxPartition >= parallelismThreshold)
{
Review Comment:
Let me think of that point and an elegant way for 1 day. I understand we
would like to satisfy the following conditions.
- We sometimes want to turn on auto reducer parallelism when `nReducers *
minPartitionFactor` is less than 1.0 but the potential reduction is not trivial
- We should keep the original behavior by default
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
