Just a couple notes about redhat and firewalls, you can run 'chkconfig --list | egrep on' and see if there are any iptables or ipchains related entries.. these can be turned off with,
'chkconfig servicename off', or, 'chkconfig --level 2 servicename off' And about node02 not being accessible, are you able to telnet to that ports 8649 from the head unit? Do you see the xml data? Its a good idea to have say the first few nodes referenced in the gmetad.conf for each cluster, this way if one is down it'll automatically fail over to the next one.. hopefully all your gmonds are able to hear each other over multicast and are picking up the same data, then they'll all be in sync and any which one can be queried by gmetad to fetch current data.
