VLSM Calculator

Allocates differently-sized subnets out of a single address block, each one sized to what it actually needs rather than forcing every subnet to the same size — the design approach behind not wasting a /24 on a point-to-point link that only needs two addresses.

The problem it solves

Fixed-size subnetting means picking one prefix length and applying it everywhere, which wastes address space fast — a 120-host office and a 2-host WAN link end up allocated identically if you're not using VLSM. Without it, a company with ten different-sized network segments might end up needing ten separate /24 blocks; with VLSM, a single, carefully planned /21 covers the same requirement and summarises far more cleanly in a routing table.

The process, step by step

• List every network and its host requirement, sorted largest to smallest — order matters here, doing it out of order wastes space
• For each one, find the smallest subnet that fits: 2ⁿ − 2 ≥ required hosts
• Allocate it starting at the next available address in the block
• The next subnet picks up right after the previous one's broadcast address
• Double-check nothing overlaps and the parent block still has enough space left for everything on the list

Where this shows up beyond the exam

VLSM needs a classless routing protocol that carries the mask with every route — OSPF, EIGRP, IS-IS, BGP all handle it; older classful protocols like RIPv1 simply can't. It's also exactly how cloud VPC design works in practice — a large /20 for a production tier, a smaller /24 for a supporting service, a /28 for a transit gateway attachment, all carved out of the same /16 VPC. Anyone who's designed an AWS or Azure network has done VLSM whether or not they called it that at the time.