[ http://issues.apache.org/jira/browse/HADOOP-150?page=comments#action_12375183 ]
Doug Cutting commented on HADOOP-150: ------------------------------------- +1 Does the job name need to be random, even? Could we allocate these sequentially too? Also, we don't even need to name the taskTracker's with a random number. We can use hostname:port, where the port is one used by one of the tasktracker's daemons (map output or umbilical). That would make these more meaningful too. Another random note: the umbilical and map output RPC servers could actually be run on the same port. The way RPC is currently implemented, all public methods on the object served are available to be called remotely. This is perhaps unsafe. Perhaps RPC should be changed so that a server names the interface whose methods calls should be restricted to... > tip and task names should reflect the job name > ---------------------------------------------- > > Key: HADOOP-150 > URL: http://issues.apache.org/jira/browse/HADOOP-150 > Project: Hadoop > Type: Improvement > Components: mapred > Versions: 0.1.1 > Reporter: Owen O'Malley > Assignee: Owen O'Malley > > The tip and task names should be related to the job id. I'd propose: > job name: job_<random 32bit/base36> > tip: tip_<job id>_[mr]_<fragment #> > task: task_<tip id>_<attempt #> > so examples would be: > job_abc123 > tip_abc123_m_00034 > task_abc123_m_00034_1 > -- 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
