The INFO message below--which appears to be an ERROR message--is found on a SAW node.
2015-02-12 18:17:58,298 INFO org.apache.hadoop.ipc.Server: IPC Server listener on 56800: readAndProcess threw exception java.io.IOException: Connection reset by peer. Count of bytes read: 0 java.io.IOException: Connection reset by peer
The tasktracker terminates the JVM after it has completed the task. In parallel, there may be a local RPC (ping) from the JVM to the tasktracker. If the tasktracker has already killed the JVM then the RPC ping will throw an exception even if task has already completed successfully.