Forum Discussion
Yes, Giorgi, I agree, that document is a bit confusing. My only guess is that they are going through a very basic example of a BIG-IP configuration. If the BIG-IP receives traffic destined for an IP address, it goes through this process to determine which Virtual Server it will use. As long as something matches, it will work. The example in that documentation is assuming there's no external routers that have specific routes for "Virtual Server Subnets". There are many people using, for example BGP on the BIG-IP, to advertise certain subnets that the BIG-IP uses for virtual servers, but the BIG-IP does not have VLAN self-ip's for. As long as you have routing set up correctly to get traffic to the BIG-IP, it will use any matching virtual servers.
You can even create virtual servers that listen for an entire subnet and have the BIG-IP do intelligent traffic routing. These are called wildcard virtual servers. You could also have a virtual server listen for traffic on 0.0.0.0/0 and if you routed all outbound traffic to the internet through the BIG-IP it could provide Firewall services with BIG-IP AFM, or Carrier Grade Source NAT'ing or source based routing.
Those are a little advanced topics, but used as an example to show you that virtual servers do not need to have a /32 address the same as a self-IP on a VLAN.
Thank James for great explanation, now it became more clear to me.