Hello, I am seeking input from anyone who has a straightforward way on how to provide the same public IP for a customer and provide the customer with two different (diverse) links. If one of the links fails, then the traffic is automatically redirects via the secondary connection.
I was considering using the check-gateway ping option, but I would prefer something that can quickly sense trouble with the link and fail over to the secondary/backup link quicker.
The primary link to the customer is a 10M/10M ethernet, the secondary link is a legacy T1 (1.5/1.5)connection. They are both delivered to the RouterBoard as ethernet (cat5e) connections. I am not interested in providing load balancing to provide the client with more speed (splitting traffic full time across both connection is what I’d like to avoid) - the T1 should only be used as backup when the primary fails.
I have address space from ARIN and I would like to give the customer a /30 or /32 public IP address that remains the same when a communication link goes bad. The customer will also be using the on-site mikrotik router as a NAT firewall. I will NOT be handing off the /30 to another router - the client would like me to handle NAT for their office.
The customer will be served from the same PoP, meaning both of the layer-2 connections back to the customer will originate from the same RB1000 unit used in our local data centre. I intend to install a routerboard at the customer (not sure which model as of yet - likely a 400 series).
I am not looking for someone to do my homework, just for a few pointers or what routing protocol I should use in this case and a couple of tips for configuration.
Many thanks 