[ http://issues.apache.org/jira/browse/HADOOP-76?page=comments#action_12445048 ] Owen O'Malley commented on HADOOP-76: -------------------------------------
The PhasedRecordWriter won't handle all of the cases, because RecordWriters can write multiple files. Furthermore, they are user code and it would be better to minimize required changes to them. A better approach would be to have a PhasedFileSystem that takes a base FileSystem and use that to commit/abort the changes. Then the framework could pass the PhasedFileSystem to the createRecordReader call and it would catch all of the files that the RecordWriter created. When the PhasedFileSystem gets a create call, it creates it in the base FileSystem with a mutated name. When the changes are commited, the files are all renamed. If the changes are aborted, the mutated filenames are deleted. > Implement speculative re-execution of reduces > --------------------------------------------- > > Key: HADOOP-76 > URL: http://issues.apache.org/jira/browse/HADOOP-76 > Project: Hadoop > Issue Type: Improvement > Components: mapred > Affects Versions: 0.1.0 > Reporter: Doug Cutting > Assigned To: Sanjay Dahiya > Priority: Minor > Attachments: Hadoop-76.patch, spec_reducev.patch > > > As a first step, reduce task outputs should go to temporary files which are > renamed when the task completes. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira
