Accordiing to http://hadoop.apache.org/common/docs/current/api/org/apache/hadoop/mapreduce/Reducer.html
the reducer copies the sorted output from each Mapper using HTTP across the network.
What port on a node is used for this data transfer? Is it 50060 by default?
It's the port of the task tracker running on each slave node, which is usually 50060 (you can check in the tasktracker log file:
2012-05-29 20:24:23,925 INFO org.apache.hadoop.http.HttpServer: listener.getLocalPort() returned 50060 webServer.getConnectors()[0].getLocalPort() returned 50060
2012-05-29 20:24:23,925 INFO org.apache.hadoop.http.HttpServer: Jetty bound to port 50060
2012-05-29 20:24:23,925 INFO org.mortbay.log: jetty-6.1.26
2012-05-29 20:24:24,283 INFO org.mortbay.log: Started [email protected]:50060
