← All lessons
Routing and network security · Lesson 7 of 7

Defence in depth

No single defence is perfect, so good networks use several layers, each one catching what gets past the last. It's called defence in depth, and it's the same idea as a castle with a moat, walls and a keep.

Segmentation splits a network into zones, so one break-in doesn't open everything. The guest Wi-Fi, the office computers, the payment machines and the servers each get their own subnet or VLAN (a virtual LAN: separate networks sharing the same switches), with a firewall between them that allows only what each zone needs.

Least privilege means every person, program and device gets only the access it needs. The receptionist's computer doesn't need to reach the database; the website's server doesn't need to start connections out to the internet. Zero trust takes it further: don't trust anything just because it's inside the network, and check every connection.

Keep systems updated, since most attacks use holes that already have fixes. Use strong sign-in: a password manager, unique passwords, and two-factor authentication, ideally passkeys or security keys. Close what isn't used: every open port and running service is one more thing to defend.

Finally, assume something will get through and plan for it. Logs and monitoring show what happened and when. Backups, kept offline or somewhere an attacker can't delete them, mean ransomware is an inconvenience instead of a disaster. And a written plan for what to do saves time when it matters.

Practice

What does defence in depth mean?

More info

Quick reference

Defence in depthSeveral layers, not one wall.
SegmentationSeparate zones (subnets, VLANs) with rules between.
VLANA separate network on shared switches.
Least privilegeOnly the access that's needed.
Zero trustCheck every connection, inside or not.
BackupsOffline or out of an attacker's reach.

The other networking courses