Nov 022011
 

Hadoop time, trying to setup a 6-nodes cluster and once it works will play with codes for some while.

  3 Responses to “Started playing with Hadoop”

  1. I have 6 machines in the cluster, s1 acts as primary name node and s2 acts as secondary name node and job tracker, all 6 machines (s1~s6) are data nodes as well (so they are running data nodes and task tracker). I guess I have everything setup but need time to confirm that.

    I think I need to have Hive or something similar.

  2. Just in case anyone encounters “Hadoop common not found” problem while running start-dfs.sh and start-mapred.sh, check HADOOP-6953, in short, need to set HADOOP_HOME in environment before running those two scripts.

  3. Downgrading to 0.20.x now – Hive does not support 0.21.x ๐Ÿ™

Sorry, the comment form is closed at this time.