AnsweredAssumed Answered

Transient datanode failures on stock hadoop?

Question asked by communityadmin on Jun 22, 2011
Latest reply on Jun 25, 2011 by lohit
In running big jobs, I see task attempts that fail with "Too many fetch-failures".  This happens on a single node, but that node doesn't seem down.  When the task runs again, it generally succeeds so this seems to be a transient load related problem.  Is this an HDFS problem?

Outcomes