We have a scenario where the customer wants the control network (PLC, HMIs) and their business network separated. Has anyone ever taken the path of installing two NICs into the server. Say for example one NIC is set to 192.168.1.200 and the other one is set to 172.16.2.200 or something else. Can the clients on the control side (192.168 range) access the gateway and launch clients at the same time as the office side (172.16)? Is there any other network administration to do? I guess the other option would be to have a router between the control network and the office network and then just forward the ports.
What are your thoughts?
Take a look at the Hirschmann Eagle20 Security Router.
Pretty nifty device.
I hope this helps.
Dennis
This is extremely common. There shouldn't be any problems or extra configuration unless you are running a cluster of Gateways. It is so common we even drew up an architecture diagram about it.
AWESOME! With a Capital ‘A’!
You mention a cluster a gateways. We have another scenario that could potentially emerge where clients would like to connect to the cluster of gateways through a router. What extra configuration would be involved here?
For clustering to work properly a client machine must be able to see both gateways. So, if the client is going through a router with a single IP and needs to see both gateways the router should have two distinct port forwards to each server. For example, port 8080 to server 1 and 8081 to server 2. If you can see both gateways through the router using the gateways IPs then there is no extra configuration.
Well, there is a little bit of extra configuration. When setting up clustering when both cluster nodes span 2 networks and when you’ll be launching clients from both sides of the network, you’ll need to enter both addresses of each server into the “HTTP Addresses” section of the clustering configuration.