I'm running hadoop job on many input files. But if one of the files is corrupted the whole job is fails.

How can I make the job to ignore the corrupted file? maybe write for me some counter/error log but not fail the whole job

Edit
Report