I could solve the problem as I figured out it was that the Hadoop filesystem was getting full.
I was monitoring the process with: /usr/bin/hadoop dfsadmin -report and noticed how the file system became full during larger processes. On 01/02/2014 06:16 PM, Law-Firms-In.com wrote: > Have you checked namenode log during this time period ? bq. Number of > data-nodes: 1 Can you deploy more data nodes ? > > =>> actually I can't I figured out now since I run all in pseudo > distributed mode. > >
