VLSM makes it possible to use different subnet masks within the same network so IP addresses are used efficiently.
Facts / key topics
- RFC: not a standalone RFC – part of the CIDR concept · OSI: Layer 3
- Requires classless routing protocols (OSPF, EIGRP, RIPv2, IS-IS, BGP)
- Not supported with classful routing (RIPv1, IGRP)
- Used in: enterprise LAN/WAN, campus, ISP
- CCNA: calculation and understanding · CCNP: design, troubleshooting, summarization side effects
Why does it exist?
Without VLSM: wasted IP addresses, all subnets must be the same size, poor scalability. Typical needs: different host requirements per segment, point-to-point links (/30, /31), loopbacks (/32).
How does it solve it?
- One larger network is split into subnets with different prefix lengths
- Design the largest subnet first; assign the smallest ones last
Step by step
- Identify host requirements per subnet
- Sort from largest → smallest
- Choose suitable prefixes (/26, /27, /30 …)
- Assign the subnets sequentially
Key terminology
| Term | Explanation |
|---|---|
| CIDR | Classless Inter-Domain Routing |
| Prefix length | /24, /27 etc. |
| Subnet boundary | Jump between subnet blocks |
| Longest Match | The most specific route wins |
Elements
| Element | Description | CCNA | CCNP |
|---|---|---|---|
| VLSM | Variable subnet masks | ✔ | ✔ |
| Route summarization | Aggregation | ✔ | ✔ |
| Discontiguous networks | Split subnets | ✖ | ✔ |
Dependencies & considerations
- The routing protocol must be classless
- Subnet calculation errors → routing faults
- Summarization can hide routes (CCNP!)
Exam tips ⚠️
- Seeing different subnet masks → that is VLSM
- Always check largest subnet first
- /30 = 2 usable hosts (the classic)
Quick summary
- Same network, different masks · efficient IP use · requires classless routing