Subnetting divides an address block into smaller networks by choosing how many leading bits identify the network, written as a prefix length or a subnet mask, so that each subnet has its own range of host addresses and its own routing entry.
Our take. Subnetting is arithmetic, and the reason candidates fear it is that courses teach it as a table to memorise rather than a calculation to perform. Learn to count in powers of two, work three examples until the pattern is obvious, and the exam questions become quick and the design questions become the interesting part.
An address has a network part and a host part, and the boundary is not fixed; you choose it. A prefix length of twenty-four means the first twenty-four bits identify the network and the remaining eight identify hosts, giving two hundred and fifty-six addresses of which two are reserved, the network address itself and the broadcast. Lengthen the prefix by one and you have two networks half the size; shorten it by one and two neighbouring blocks merge into one. That is the whole mechanism. Everything else, the masks written in dotted form, the tables of block sizes, the variable-length designs where subnets of different sizes sit side by side, is that one operation applied with care.
The reason it matters beyond the exam is that a subnet is a boundary. Devices in the same subnet reach each other directly, through a switch; devices in different subnets need a router between them, which is a place to filter, to log and to contain. Broadcast traffic stops at the subnet's edge, so a subnet that is too large is noisy and a subnet that is too small runs out of addresses. Security design uses the boundary deliberately: servers in one subnet, users in another, the building's cameras in a third, with a firewall rule set between them that only exists because the subnets do. Subnetting is where segmentation starts.
Prefix length
Addresses in the block
Usable for hosts
Typical use
/30
4
2
A point-to-point link between two routers
/29
8
6
A tiny DMZ or a handful of servers
/27
32
30
A small office segment
/24
256
254
The default everyone learns first
/22
1,024
1,022
A large office floor or a campus segment
/16
65,536
65,534
A whole site, split further inside
In practice
Count the host bits. A /26 leaves six host bits; two to the sixth is sixty-four addresses per subnet.
Find the block boundaries. Sixty-four-address subnets start at 0, 64, 128 and 192 in the last octet.
Place the address. A host with 130 in the last octet falls in the block starting at 128; that block runs to 191.
Reserve the ends. 128 is the network address, 191 the broadcast, 129 to 190 are usable, and 130 is a valid host. That is the entire exam question, and it takes fifteen seconds once the pattern is familiar.
A company is handed a /22 by its provider and needs a server segment, a user segment, a guest segment and a management segment. Splitting it into four /24s is the obvious design and the wrong one, because the server segment needs thirty addresses and the user segment needs six hundred. Variable-length subnetting gives the users a /23, the servers a /26, the guests a /25 and management a /27, with room left over, and each boundary becomes a firewall rule the flat design could never have had.
Addressing is the numbering scheme; subnetting is the decision about where each address's network part ends. You cannot subnet without understanding the address, and you can understand the address without ever subnetting.
A VLAN is a switch feature that separates devices at layer two; a subnet is a layer-three boundary. They are usually paired one to one, which is why people confuse them, and they are different mechanisms doing different halves of the same separation.
Routing moves packets between subnets. Subnetting creates the boundaries routing has to cross, and a routing table is a list of subnets and where to send traffic for each.
Key takeaways
→Decide how many leading bits belong to the network; that one choice is all of subnetting.
→A subnet is a boundary: direct reach inside it, a router and a filter between them.
→Count host bits, find block boundaries, reserve the ends; practise until it is automatic.
Certifications that test this
Vendor exams whose syllabus covers this concept — facts, cost and a preparation path on each page.
A rotating selection from the course directory, drawn from the subcategories where this concept is taught rather than picked for it. Details, price and the provider link are on the course page.
.NET is a modern, innovative, open source development platform and developers love it. NET best suits businesses that l…
Udemy
FAQ
Why does every networking exam test subnetting?
Because it is the arithmetic underneath every design and every troubleshooting step, and because it is quick to examine. A candidate who cannot say which subnet an address belongs to cannot read a routing table or a firewall rule, and the exams know it.
What is CIDR?
Classless inter-domain routing: the scheme that replaced fixed address classes with prefix lengths of any size, written with a slash. Every prefix length on this page is CIDR notation, and the classes it replaced survive only in old textbooks and one exam question.
How small should a subnet be?
As small as its purpose allows, because smaller subnets contain broadcast traffic and give you more boundaries to filter at. Leave room for growth on the segments that grow, such as users, and keep infrastructure segments tight.
Sources
The primary text this definition rests on. Read it before you trust ours.
