VLSM using Class A/B subnet masks

Pupil

Is this something I should be expecting on the exams? Are there questions that will ask me to setup a network with like say 500 hosts? 10000 hosts?

I'm comfortable enough doing VLSM with Class C subnet masks, but it's tricky with subnet masks that provide huge number of hosts like /9 to /15.

Is this something done on real networks? Who would want a broadcast domain that large?

xnx

Don't have time to explain but here's some info from subnetting.net


Question Type 3: You have been asked to create a subnet mask for the 172.16.0.0 network. Your organization requires 900 subnets, with at least 50 hosts per subnet. What subnet mask should you use?
Step 1: Determine how many subnet bits (1's), you have to add to the classful boundary to cover the number of required subnets.

• The IP address given was a class B address, making the first 16 subnet bits static.
• Using the **** sheet, find the exponent of 2 that is equal to or greater than the number of subnets we require (900). We can quickly see that 10 additional subnet bits will give us 1,024 subnets. Make note of the corresponding subnet mask. In this case, 255.255.255.192. The third octet is eight 1's, and the four is two 1's. We can count up from the bottom on our **** sheet to get to 192.

Step 2: Confirm the number of remaining 0's will cover our required hosts. In this case, there are 6 remaining 0's - 2^6-2=62, which is more than enough for our host requirements.
Our subnet mask is 255.255.255.192. Giving us 1024 subnets and 62 hosts per subnet.

rocdamike

xnx explained it well. Here is another example of how to know which subnet to use for a given number of hosts (number of subnets not specified). I have just changed the question slightly:

Question Type 3: You have been asked to create a subnet mask for the 10.0.0.0 network. Your organization requires 500 hosts. What subnet mask should you use?

1. Use the formula 2^y -2 = no. of hosts
2⁹ -2 = 510 hosts (this covers the 500 hosts that we need)

2. No. of host bits in default Class A network = 24

3. Therefore total number of subnet bits we need = 24 - 9 = 15

4. Subnet mask we need is therefore 255.255.254.0 (add 15 masked bits to the default class A mask)

xnx

Just to clarify that what I posted is copied off subnetting.net... HOWEVER it's the exact same method I use and would recommend

Pupil

You may have misunderstood my question. I was asking about VLSM not general CIDR subnetting. Here's a question that I had in mind:

Using the network address 172.20.0.0/16, subnet to meet the following requirements:
- one subnet to support 1000 hosts
- two subnets to support 120 hosts
- 3 subnets to support 25 hosts
- 5 subnets for WAN links

Is this something I could expect on the exams or are VLSM questions purely only done using Class C addresses?

devils_haircut

Pupil wrote: »

Is this something I could expect on the exams or are VLSM questions purely only done using Class C addresses?

You could it expect it on the exam, but I'm not sure why it would matter. Calculating VLSM for networks is the same regardless of the size of the block.

I'd also like to mention that it's a bit confusing to be mixing the term "Class" into questions regarding VLSM. VLSM only exists in classless networks, not classful.


EDIT: Another TE thread that covers VLSM, CIDR, and classful/classless topics (fairly in depth)

http://www.techexams.net/forums/ccna-ccent/15813-classeless-vs-vlsm-vs-cidr.html

Jon_Cisco

I would expect to see questions with 500 hosts but probably not much higher.
They are not writing math questions just to stump you.

Good Luck!