We have two node cluster, one master and one slave.
step1 ) stop hadoop ( /bin/stop-all.sh ) step2 ) modify content of /conf/master file from "master" to "slave" step3) start hadoop ( /bin/start-all.sh ) Are the above steps sufficient for moving "Secondary NameNode" daemon process from "master" on to "slave" machine ? Thanks, Giridhar. From: Giridhar Addepalli [mailto:[email protected]] Sent: Friday, May 13, 2011 2:43 PM To: [email protected] Subject: Moving Secondary NameNode Hi, As of now, primary namenode and secondary namenode are running on the same machine in our configuration. As both are RAM heavy processes, we want to move secondary namenode to another machine in the cluster. What does this move take? Please refer me to some article which discusses these issues (or) please give me the steps to follow. Thanks, Giridhar.
