apparent routing problem

All, 

I am in the process of setting up a six node piranha load balancing
cluster: a main load balancer, a backup load balancer, and four web
servers. I'm using RedHat Advanced server version 2.1, kernel 2.4.9-e.3,
piranha version 0.7.0-3. I'm kind of green when it comes to networking,
so please bear with me.

The cluster appears to be up and running as I can see the lvs daemon and
the four nanny processes running. Failover seems to work fine also. The
problem is that I cannot seem to hit the web servers from the internet.
I can, however, hit them from one 
server on my internal network. 

The private ip addresses for the load balancers are 10.0.1.6 (primary)
and 10.0.1.7 (backup). The NAT ip address is 10.0.1.254, and I have set
this up as the default route for the four real servers. 

The public ip addresses for the load balancers are 10.0.3.133 (primary)
and 10.0.3.134 (backup). The virtual server ip address is 10.0.3.132,
and I have a real public ip address, 216.x.x.x, mapped to it in the
firewall. I have tried by passing the firewall 
completely by changing the virtual server address to 216.x.x.x and
plugging the load balancers directly to the internet. The same symptoms
showed. 

On both routers, in the /etc/sysctl.conf file, I have the following
lines: 
net.ipv4.ip_forward = 1 
net.ipv4.conf.default.rp_filter = 1 
kernel.sysrq = 0 

I tried using ipchains, but apparently it's not been built into the
kernel. I get a 'protocol not supported' message when I try to run the
following line: 

ipchains -A forward -j MASQ -s 10.0.1.0/24 -d 0.0.0.0 

What am I doing wrong? Why is it that I can't see the web servers. I
suspect that I'm supposed to be doing something with iptables.
from the internet? 

Thank you. 

-Stephen Spalding