Hi,
After some further testing, i can perhaps give some more detailed techn.
info on my problem:
ganglia master: linux rh. works ok, web interface & local gmond client
1 node i want to monitor is a solaris 9 x86 machine which is multihomed.
the idea is that all ganglia traffic goes via the management network,
which is interface bge1 on the solaris.
this is ifconfig -a on the solaris:
bash-2.05# ifconfig -a
lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
bge0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index
2
inet 193.110.158.170 netmask ffffffe0 broadcast 193.110.158.191
ether 0:9:3d:0:22:f9
bge1: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index
3
inet 192.168.2.170 netmask ffffff00 broadcast 192.168.2.255
ether 0:9:3d:0:24:eb
this is my current configuration on the solaris /etc/gmond.conf
concerning the network parameters:
/* Feel free to specify as many udp_send_channels as you like. Gmond
used to only support having a single channel */
udp_send_channel {
mcast_join = 239.2.11.71
mcast_if = bge1
port = 8649
}
/* You can specify as many udp_recv_channels as you like as well. */
udp_recv_channel {
mcast_join = 239.2.11.71
mcast_if = bge1
port = 8649
bind = 239.2.11.71
However, no matter what i do, netstat -an on the solaris reports this:
bash-2.05# netstat -an | grep 8649
*.8649 Idle
193.110.158.170.43867 239.2.11.71.8649 Connected
*.8649 *.* 0 0 65536 0
LISTEN
so it still seems to bind to the wrong interface.
if i change bind = 239.2.11.71 to bind = 192.168.2.170, nothing changes
in the netstat output.
I also added a route that says that traffic to host 239.2.11.71 should
go to 192.168.2.170 (the address of intf bge1), but still no luck.
What more should i do to have the solaris machine appear in the web
interface ? I'm totally lost here.
--
Tom Van Overbeke <[EMAIL PROTECTED]>
