[
https://issues.apache.org/jira/browse/HIVE-9251?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14263821#comment-14263821
]
Rui Li commented on HIVE-9251:
------------------------------
Basically the following problems will lead to small number of reducers:
# At start-up, it may take some time for executors to register with driver.
Therefore we may get inaccurate # of executors.
# We rely on {{spark.executor.cores}} to get # of cores per executor, which is
not available to standalone mode. Therefore, # of total cores will be the same
as # of executors in standalone mode.
# We didn't consider the maximum data size a reducer can handle.
To solve the first two problems, we may need to ask spark to expose more
information about granted resources to user.
> SetSparkReducerParallelism is likely to set too small number of reducers
> [Spark Branch]
> ---------------------------------------------------------------------------------------
>
> Key: HIVE-9251
> URL: https://issues.apache.org/jira/browse/HIVE-9251
> Project: Hive
> Issue Type: Sub-task
> Components: Spark
> Reporter: Rui Li
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
