[ http://issues.apache.org/jira/browse/HADOOP-577?page=comments#action_12440187 ] Sanjay Dahiya commented on HADOOP-577: --------------------------------------
Does it make sense to use JobTracker history mechanism ( HADOOP-239) to include this information as well ? I think it does. > Job tracker should collect detailed information about tasks and return it to > the client > --------------------------------------------------------------------------------------- > > Key: HADOOP-577 > URL: http://issues.apache.org/jira/browse/HADOOP-577 > Project: Hadoop > Issue Type: Bug > Components: mapred > Environment: all > Reporter: arkady borkovsky > > Mapreduce driver should get detailed information about task execution -- at > least at the end of the job so that it can be repoted to the user (stored in > a "normal" or HDFS file, printed to the stdout, whatever). > This information should include, at least: > -- task execution start and end > -- the number of record read and written > -- the number of butes read and written > for each task (for each attempt) > It would also be nice to have this infromation while the job is running, > updated at least when a task is started or ended. > However, this is just "nice to have", while the final report is quite > essential. -- 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
