Monitoring Data - Big Data Made Easy: A Working Guide to the Complete Hadoop Toolset - page 208

In the same line, I named the cluster “my cluster” for purposes of Ganglia data collection. On the collector server
hc1nn, I set up the gmond.conf file using this name and the other parameters that follow, like this:
[root@hc1nn etc]# vi /etc/ganglia/gmond.conf
cluster {
name = "my cluster"
owner = "unspecified"
latlong = "unspecified"
url = "unspecified"
}
udp_send_channel {
mcast_join = hc1nn
port = 8649
}
udp_recv_channel {
port = 8649
}
tcp_accept_channel {
port = 8649
}
I then restart the collector and node daemons on the collector server hc1nn by using the Linux service
command:
service gmetad restart
service gmond restart
On each data node to be monitored (hc1r1m1, hc1r1m2, hc1r1m3), I set up the configuration of the Ganglia
gmond process by editing the gmond.conf file under /etc/ganglia:
[root@hc1r1m1 etc]# vi /etc/ganglia/gmond.conf
cluster {
name = "my cluster"
owner = "unspecified"
latlong = "unspecified"
url = "unspecified"
}
udp_send_channel {
mcast_join = hc1nn
port = 8649
}