Hi, for the purpose of a monitoring project, I am trying to find out how to programmaticallyeally from the client interface (JobClient etc.), do the following in Hadoop MR:
1) Given a taskID (or taskattemptID), figure out the ID of the node/server (IP address or hostname) where the map/reduce task is running. 2) From the map method of my mapper class (overloading Mapper), get a hook to the Partitioner class. Essentially, what I want to do here is to count - as the map task progresses - the output bytes of a map task that will be fetched/processed by each reducer task. *Please note I am using Hadoop 1.0.4* Any help or hint would be greatly appreciated! Thanks
