[MAPREDUCE-5049] CombineFileInputFormat counts all compressed files non-splitable

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.1.1
• Fix Version/s: 1.2.0
• Hadoop Flags: Reviewed
• Assignee: Sandy Ryza
• Reporter: Sandy Ryza
• Created: 06/Mar/13 03:52
• Updated: 15/May/13 05:16
• Resolved: 07/Mar/13 22:34

Description

In branch-1, CombineFileInputFormat doesn't take SplittableCompressionCodec into account and thinks that all compressible input files aren't splittable. This is a regression from when handling for non-splitable compression codecs was originally added in MAPREDUCE-1597, and seems to have somehow gotten in when the code was pulled from 0.22 to branch-1.