Bytes for Internal network addresses

Question

I have a repeating question in test in my college, which bothers me all the time.

I am translating it from another language, so bear with me:

Internal network addresses (please enter number of bytes for each field, use * if all bytes)

A class: __ __ __ __

B class: __ __ __ __

C class: __ __ __ __

I know that it should look like this:

A: ? * * * B: ? ? * * ...

I just not sure what exactly they are asking from me, and what kind of math is needed there.

Answer 1

I think the question is asking about Classful Networks - something which has largely been confined to history due to CIDR (Classless Inter-Domain Routing).

TL;DR

The question isn't clear, appears to contain fundamental misunderstanding, and uses an archaic "fact" (which is commonly confused) for grading students. I'd push back.

---

As they are asking about "internal network addresses", then it's possible they are referring to the reserved address spaces for Private Network Addresses (however I don't think this is their intention):

• 10.0.0.0/8 - one Class A network
• 172.16.0.0/12 - 16x Class B networks, back to back
• 192.168.0.0/16 - 256x Class C networks

---

The question goes on to say "please enter number of bytes for each field"... An IPv4 address is a 32-bit address, that is broken into four 8-bit (1 byte / octet) chunks and written in decimal. Let's presume that by "... number of bytes ..." they actually mean "... number of bits ..." - especially as there are four "fields" given.

He laughed and said, yes, bytes

---

In this case, the question could probably be better worded as below (though it's still an odd question to ask):

Write the number of bits of each octet that is used to represent the network prefix.

Huh... "network prefix"?

That /8, /12 or /16 at the end of the reserved address spaces above is one method of representing a subnet mask. Another representation that you may be more familiar with is a value such as 255.255.255.0 (which equates to /24).

• The /x notation states "x bits are used for the network address"
• The 255.255.255.0 notation states "this is the mask used for the network address".

255.255.255.0 or /24 can be written out in full, in binary as:

dec      255      255      255        0  
bin   11111111 11111111 11111111 00000000

Class A

So, given that a Class A network is /8 (or 255.0.0.0), it could be written as:

dec      255        0        0        0  
bin   11111111 00000000 00000000 00000000

The answer I'd give would be: 8, 0, 0, 0.

Class B

So, given that a Class B network is /16 (or 255.255.0.0), it could be written as:

dec      255      255       0        0  
bin   11111111 11111111 00000000 00000000

The answer I'd give would be: 8, 8, 0, 0.

Class C

So, given that a Class C network is /24 (or 255.255.255.0), it could be written as:

dec      255      255      255        0  
bin   11111111 11111111 11111111 00000000

The answer I'd give would be: 8, 8, 8, 0.

---

Conversely, if what they actually wanted to ask was:

Write the number of bits of each octet that is used to represent the host address.

I'd use the following responses - note the bits are inverted from the subnet mask to produce the "host address mask".

Class A

So, given that a Class A network is /8 (or 255.0.0.0), it could be written as:

dec      255        0        0        0  
bin   11111111 00000000 00000000 00000000  <== subnet mask
bin   00000000 11111111 11111111 11111111  <== host address mask

The answer I'd give would be: 0, 8, 8, 8.

Class B

So, given that a Class B network is /16 (or 255.255.0.0), it could be written as:

dec      255      255        0        0  
bin   11111111 11111111 00000000 00000000
bin   00000000 00000000 11111111 11111111  <== host address mask

The answer I'd give would be: 0, 0, 8, 8.

Class C

So, given that a Class C network is /24 (or 255.255.255.0), it could be written as:

dec      255      255      255        0  
bin   11111111 11111111 11111111 00000000
bin   00000000 00000000 00000000 11111111  <== host address mask

The answer I'd give would be: 0, 0, 0, 8.