You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The current Hadoop platform has at least one issue when being used by an MR2 application; it winds up configuring the job with a single reduce task, due to https://issues.apache.org/jira/browse/MAPREDUCE-4288. So what we'd like is a new platform (which extends the HadoopPlatform). This should perhaps by default set the number of reducers to one per node, as that's information we should be able to get back from YarnClusterMetrics.
The text was updated successfully, but these errors were encountered:
The current Hadoop platform has at least one issue when being used by an MR2 application; it winds up configuring the job with a single reduce task, due to https://issues.apache.org/jira/browse/MAPREDUCE-4288. So what we'd like is a new platform (which extends the HadoopPlatform). This should perhaps by default set the number of reducers to one per node, as that's information we should be able to get back from YarnClusterMetrics.
The text was updated successfully, but these errors were encountered: