VLSM and CIDR Notation

Variable Length Subnet Masking (VLSM) is the practice of using different subnet masks for different subnets within the same major network (i.e., the same classful address block). Classless Inter-Domain Routing (CIDR) is the method for allocating IP addresses and routing IP packets without regard to the traditional classful boundaries (A, B, C). CIDR uses prefix notation (e.g., /24) to denote the network portion of an address. VLSM is essentially an application of CIDR: it allows a network to be subnetted into subnets of varying sizes, reducing address waste.

Before CIDR (RFC 1518, 1519), IP addresses were assigned based on class: Class A (/8), Class B (/16), Class C (/24). This led to massive waste because organizations rarely used all addresses in their allocated block. For example, a company needing 300 addresses would receive a Class B (65,534 addresses), wasting over 65,000 IPs. By the early 1990s, the Internet was running out of addresses and routing tables were exploding. CIDR solved both problems: it allowed ISPs to allocate address blocks in any size (e.g., a /23 for 510 addresses) and enabled route aggregation (supernetting) to keep routing tables manageable.

VLSM works by allowing a subnet to be further subdivided into smaller subnets, each with its own subnet mask. The key is that the subnet mask must be extended (i.e., more bits set to 1) for the smaller subnets. For example, a /24 network (255.255.255.0) can be divided into two /25 subnets (each with 126 usable hosts), or four /26 subnets (62 hosts each), or one /25 and two /26, etc. The routing protocol must support VLSM (i.e., it must carry the subnet mask in its updates). Classful routing protocols like RIPv1 and IGRP do not support VLSM; classless protocols like RIPv2, OSPF, EIGRP, and BGP do.

Traditional subnetting uses a fixed-length subnet mask (FLSM) across all subnets. For example, a /24 network subnetted into /26 subnets gives four equal-sized subnets. If one subnet needs only 10 hosts and another needs 50, FLSM forces the 10-host subnet to use a /26 (62 hosts), wasting 52 addresses. VLSM allows the 10-host subnet to use a /28 (14 hosts) and the 50-host subnet to use a /26, reducing waste.

CIDR notation is written as an IP address followed by a slash and the number of bits in the network prefix. For example, 192.168.1.0/24 means the first 24 bits are the network portion. The subnet mask in dotted decimal is 255.255.255.0. Common prefix lengths: - /8 (255.0.0.0) - 16,777,214 hosts - /16 (255.255.0.0) - 65,534 hosts - /24 (255.255.255.0) - 254 hosts - /30 (255.255.255.252) - 2 hosts (point-to-point links) - /32 (255.255.255.255) - 1 host (loopback or host route)

CIDR enables route aggregation, also called supernetting or summarization. Instead of advertising multiple contiguous smaller networks, a router can advertise a single larger network that encompasses them. For example, instead of advertising 192.168.0.0/24, 192.168.1.0/24, 192.168.2.0/24, and 192.168.3.0/24 separately, a router can advertise 192.168.0.0/22 (which covers all four /24s). This reduces routing table size and improves network stability.

To perform VLSM, follow these steps: 1. List all subnets in descending order of host requirements. 2. Determine the smallest subnet mask that accommodates the largest subnet (including network and broadcast addresses). 3. Assign the first available network address to that subnet. 4. Repeat for the next largest subnet, using the next available address block.

Example: You have network 10.0.0.0/8 and need subnets with 500, 200, 50, and 2 hosts. - 500 hosts: need /23 (510 hosts). Subnet: 10.0.0.0/23 (range 10.0.0.0 - 10.0.1.255). - 200 hosts: need /24 (254 hosts). Next available: 10.0.2.0/24. - 50 hosts: need /26 (62 hosts). Next available: 10.0.3.0/26 (range 10.0.3.0 - 10.0.3.63). - 2 hosts: need /30 (2 hosts). Next available: 10.0.3.64/30 (range 10.0.3.64 - 10.0.3.67). Note that the /26 and /30 are inside the /24 space; VLSM allows this nesting.

Support for VLSM requires classless routing protocols. RIPv2, OSPF, EIGRP, IS-IS, and BGP all include the subnet mask in their route advertisements. RIPv1 and IGRP do not; they assume the default classful mask. When using VLSM, you must ensure all routers run a classless protocol or manually configure static routes with masks. On the exam, remember that OSPF and EIGRP are classless by default; RIPv2 is classless; RIPv1 is classful.

VLSM can lead to discontiguous subnets, where subnets of the same major network are separated by another major network. For example, 10.1.0.0/16 and 10.2.0.0/16 are both subnets of 10.0.0.0/8 but are separated by a different network. Classful routing protocols will automatically summarize to the classful boundary (10.0.0.0/8) and will not see the individual subnets, causing routing problems. Classless protocols can handle discontiguous networks if summarization is disabled. The exam tests this: discontiguous networks cause routing loops with classful protocols.

CIDR is the foundation of Internet routing. ISPs are allocated blocks of IP addresses from Regional Internet Registries (RIRs) in CIDR notation. For example, ARIN may allocate 198.51.100.0/24 to an ISP. The ISP can then subdivide it into smaller blocks for customers. BGP routers exchange routes using CIDR prefixes. Without CIDR, the global routing table would be hundreds of times larger.

On a Cisco router, to configure a static route with a /25 mask:

ip route 192.168.1.128 255.255.255.128 10.0.0.1

Or using CIDR notation:

ip route 192.168.1.128/25 10.0.0.1

To enable RIPv2 (classless):

router rip
 version 2
 network 10.0.0.0
 no auto-summary

To verify routing table and see masks:

show ip route

The output includes the prefix length, e.g., 10.0.0.0/8 or 10.0.1.0/24.

Memorize these: - /30: 2 usable hosts (point-to-point) - /29: 6 usable hosts - /27: 30 usable hosts - /26: 62 usable hosts - /25: 126 usable hosts - /24: 254 usable hosts - /23: 510 usable hosts - /22: 1022 usable hosts - /21: 2046 usable hosts - /20: 4094 usable hosts

Given network 172.16.0.0/16, create subnets for: 1000 hosts, 500 hosts, 250 hosts, 100 hosts, 2 hosts (WAN link).

Notice that the /25 and /30 are nested within the /24 space; this is VLSM.

A university has a single /16 network (e.g., 10.0.0.0/16) and needs to allocate subnets to different departments: Engineering (2000 hosts), Business (500 hosts), Admin (100 hosts), and dormitories (4000 hosts). Using VLSM, the network engineer assigns a /20 to dorms (4094 hosts), a /21 to Engineering (2046 hosts), a /23 to Business (510 hosts), and a /25 to Admin (126 hosts). This saves thousands of addresses compared to using a fixed /20 for all. The routing protocol is OSPF, which supports VLSM. Without VLSM, the engineer would have to waste addresses or use multiple major networks. Performance is fine as long as route summarization is configured at the distribution layer (e.g., summarizing all /20s into a single /16). Misconfiguration: if auto-summary is enabled on a router running RIPv2, it would summarize to 10.0.0.0/8 and lose visibility of individual subnets, causing reachability issues.

A data center uses /30 subnets for all inter-switch and inter-router links. For 50 point-to-point links, only 100 IP addresses are needed (50 networks x 2 usable hosts each). Using VLSM, these /30s are carved out of a larger block, say 10.10.0.0/24. The engineer assigns 10.10.0.0/30, 10.10.0.4/30, etc., up to 10.10.0.196/30. This uses only 200 IPs out of 256 in the /24, leaving room for future links. If FLSM were used with /24, each link would waste 252 IPs. Common mistake: using /29 for point-to-point links wastes 6 IPs per link, which adds up. The engineer must ensure that loopback interfaces use /32 masks. BGP is used for external routing; BGP relies on CIDR for route aggregation.

An ISP receives a /22 block (1024 addresses) from its upstream provider. It needs to allocate subnets to customers: one customer needs 300 hosts, another needs 50, and several need 2-host links. Using VLSM, the ISP assigns a /23 (510 hosts) to the 300-host customer, a /26 (62 hosts) to the 50-host customer, and /30s for point-to-point links. The ISP must ensure that the subnets do not overlap and that the customer routers use classless routing (e.g., BGP or static routes). The ISP also advertises the /22 as a single prefix to the Internet, aggregating all customer subnets. If the ISP misconfigures and advertises individual /24s, the global routing table grows unnecessarily. Performance consideration: route aggregation reduces BGP table size and convergence time.