Very new to network engineering. I just watched a really good series of videos on YouTube explaining IP addresses, subnet masks and the class system for IPv4.

There was one bit which got me confused however.

In lesson 2 (https://www.youtube.com/watch?v=XQ3T14SIlV4), where he goes through subnet masking, he gives an example where he uses an IP address with subnet mask This is at around 4:51.

However, in the subsequent lesson 3 (https://youtu.be/4xlzlgYGqW8), where he teaches default subnet masks, he says class C IP addresses must use subnet masks of at least

Does this mean the example used in lesson 2 was false? Or is it possible to use subnet masks smaller than the default values for a given class?

Either way he got the point across very well, just want to make sure I understand everything.

Thanks all.

  • 3
    Network classes died in 1993 with the introduction of CIDR. Please let them rest in peace. Accordingly, there is no "default subnet mask" unless you define one for your network. – Zac67 Apr 30 at 12:05

Classful addressing is a relic of history. Today, we use Variable Length Subnet Masks (VLSM) and that's why you need to specify a netmask when configuring addresses and routes.

You'll also see the term Classless Inter-Domain Routing (CIDR) which is the same concept -- configurable subnet sizes independent of the address number -- applied to routing information exchanged among different (usually BGP-speaking) networks.

When you see educational resources referring to address classes, honestly, this is because the material is bad, and more than twenty years out of date. The exception to this are the "Class D" and "Class E" parts of the IPv4 address space, which are used for Multicast, and unused, respectively.

| improve this answer | |
  • 1
    Thanks Jeff, will check that out. Out of interest, would that have been an invalid IP address and subnet mask combo back in the day (20 years ago) when there was classful addressing? – George Wilson Apr 30 at 12:42
  • 1
    @GeorgeWilson would have belonged to C class, which you could have subnetted but afaik not supernetted - then (or /21) would in fact have been illegal. But that logic is all long obsolete, thanks to CIDR and VLSM. – Zac67 Apr 30 at 14:30

Not the answer you're looking for? Browse other questions tagged or ask your own question.