Moving this to the user list. Please use user lists for such queries and dev lists only for development discussions about Hadoop MapReduce alone.
2011/3/5 周俊清 <[email protected]>: > Hi,everyone > I am using apache version of hadoop0.20.2 to do my job. now i want to add a > new machine to my cluster as TaskTracker and DataNode. How can i do that > without restart the whole hadoop system? or just run some commands on the new > machine? any help will be appreciated.thanks On your master machine, the new slave's hostname to your slaves file and reissue a start-all.sh (no stop-all.sh is required.) Or ssh into your slave and issue "hadoop-daemon.sh start tasktracker" [similarly, for datanode]. -- Harsh J www.harshj.com
