VLSM support vs. classful

johnifanx98

Is it fair to say the major advantage of VLSM is to support discontiguous networks?

nkillgore

A lot of times it is used to divide a limited address space into smaller chunks.

So, every router to router connection in our company is on a /30 network. (e.g. 10.1.1.0 - 10.1.1.3, with the routers being at 10.1.1.1 and 10.1.1.2).

If you only had 10.1.1.0/24 to work with, you could put your hosts in 10.1.1.128/25 and divide the lower half into other subnets for servers(10.1.1.0/26), network management(10.1.1.64/28 ), etc.

I'm not sure about extremely large organizations, but most places aren't going to be large enough to have to do this with an RFC1918 address. The place I have seen it used most is when we connect to the ISP via a publicly routed IPv4 address. They want to use as few of those per connection as they possibly can. You might also do this on externally facing hosts if you only had a few addresses and needed to divide them up.