[HADOOP-1029] streaming doesn't work with multiple maps

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 0.11.2
• Fix Version/s: 0.12.0
• Assignee: Arun Murthy
• Reporter: Arun Murthy
• Created: 22/Feb/07 03:26
• Updated: 08/Jul/09 17:05
• Resolved: 23/Feb/07 17:57

Description

StreamInputFormat doesn't 'seek' to the right offset for the split, therefore leading to a situation where every map reads the same input-split and processes it; also the StreamLineRecordReader has a bug where it doesn't clear out the contents of previous keys/values and that data is appended to each subsequent key/value.

Issue Links

• blocks: HADOOP-878 reducer NONE does not work with multiple maps