I think I understand now. This configuration will require that I have installed Nginx in all servers listed below:

upstream proxy {
server 1.localserverip:8000;
server 2.localserverip:8000;
server 3.localserverip:8000;
server 4.localserverip:8000;
}

when in real life, I will have Nginx installed into 1.localserverip only.
Presuming that I plan to add another balancer to the equation:

upstream proxy {
server 1.localserverip:8000;
server 2.localserverip:8000;
}

while I keep serving the php files from all servers:

upstream php {
server 1.localserverip:9000;
server 2.localserverip:9000;
server 3.localserverip:9000;
server 4.localserverip:9000;
}

That will require I install Nginx into 1.localserverip and 2.localserverip. So far, it is simple and logical.
My concern is what will happen with my fastcgi as well the try_files directives:

upstream proxy {
server 1.localserverip:8000;
server 2.localserverip:8000;
}

upstream php {
server 1.localserverip:9000;
server 2.localserverip:9000;
server 3.localserverip:9000;
server 4.localserverip:9000;
}

location / {
proxy_pass http://proxy;
}

location /forum/ {
try_files $uri $uri/ /forum/data.php$args;
}

location ~ \.php$ {
fastcgi_pass php;
include fastcgi.conf;
}

Can you post a quick configuration that will work with the above setup? Thanks.