
Casey M. answered 03/22/16
Tutor
New to Wyzant
Grad student specializing in Python programming
What you are referring to relates to a topic in computer networking known as subnetting. This is essentially when a given block of addresses (e.g. 135.50.20.0/22) is divided into several other blocks of addresses (e.g. 135.50.20.0/24, 135.50.22.0/24, etc.). Essentially, that last bit, the /22 and /24 indicate the bits which refer to the network prefix, meaning the remaining bits refer to the host number. The network prefix is the set of leading bits which all addresses on given network share. The host number is the set of trailing bits which distinguish hosts on a given network from each other.
To make this clearer, it will be helpful to write out the address 135.50.20.0 in binary format (from here on out I'm assuming you understand binary as opposed to decimal format; let me know if that is not the case and I will provide more information):
10000111.00110010.00010100.00000000
Note that each individual number in an IPv4 (as opposed to IPv6) address is an 8-bit number, meaning they can range in value from 0-255, giving a total of 256 different combinations.
With that groundwork laid, we consider the address space that the ISP has 135.50.20.0/22. We write that out in binary format, with the bold bits indicating the network prefix:
10000111.00110010.00010100.00000000
Notice that the first 22 bits are emboldened (/22). This means that the ISP has the address space consisting of all addresses prefixed by the bits 10000111.00110010.000101. Thus, the final 10 bits (8*4 = 32; 32-22 = 10) of the addresses in the address space are variable. If the last 10 bits are variable, then there are 210 = 1024 different possible combinations of those trailing 10 bits, that is there are 1024 different addresses in the ISP's address space.
We know that the subnets 135.50.20.0/24 & 135.50.22.0/24 are already assigned. Using the same logic as above, the /24 indicates a subnet of 28 = 256 (32-24 = 8) addresses. Thus, we subtract 512 (256+256 = 512) from 1024, leaving 512 addresses unassigned in the ISP's address space.
So, subnets can only be allocated in powers of two since it all revolved around binary logic. Thus, the closest we can get to a subnet of 100 is a subnet of 128. Since 512 addresses remain, we divide them into groups of 128. 512/128 = 4, so the remaining addresses can be divided into 4 additional subnets of 128 addresses, so 4 additional customers can be serviced with at least 100 each. This gives a total of 6 customers served by the ISP, with the first two having the larger address spaces of 256 addresses each, and the last 4 having 128 each.
For further study, below are some helpful links:
https://en.wikipedia.org/wiki/Subnetwork
https://www.aelius.com/njh/subnet_sheet.html
http://www.webopedia.com/TERM/S/subnet_mask.html