Typical, I struggle with this for a day, and then five minutes after posting here, I figure it out...
So for anyone else experiencing this problem:
You have to set the --host option on the run.sh script. E.g.:
./run.sh -c all --host=192.168.2.2
Hope that helps someone!
I got the same problem and I don't want to modify the startup script, so I worked more till I found the origin of error.
The problem is in binding the ANY_ADDR 0.0.0.0 to multicast socket on a linux kernel with ipv6 enabled.
If you disable ipv6 from kernel the binding of multicast is ok and the services related to 'all' configuration runs smooth.
HI Thanks for posting this, I have another question how does the other server know to replicate the session info.
I have it set it as HTTP Session Replication. When both servers come up they think they are only 1 member in the cluster.
Thanks Ric it helped me!
rather than turn off in kernel you may want to use this.