Jan 15, 2011 11:23 PM

Like (0)
Hi all,
This has likely been covered but I can't seem to find the reason behind the answer to this question.
You are working on a network that cannot interpret classless addressing. Your network ID is 145.27.0.0. What it the THEORETICAL max. subnets you can create on this network?
Answer is apparently: 254
WHAT??. This is a classeless network, subnetting it would not work therefore you can't have any subnets. Only networks that can interpret classless networking schemes can use subnetting. Therefore the answer is 0. You can have 1 network but no subnets.
BUT, if it the question is asking for THEORETICAL max. number of subnets because you are about to change the network devices to recognise subnetting, the the answer should be 16382 subnets with 2 hosts per network. or 145.27.0.0/30
What have I missed in order to not get the answer of 254?
I would appreciate any help on the reasoning behind the answer given.
Kindest regards,
Bouts
Aloha Bouts,
Let me see if I can help you out. Since your network cannot understand "classless" then it means the network is "classful". Classful pertains to the default Class A, B, and C networks. You can still subnet these "classful" networks, but the subnets must be the same throughout.
Now taking the Class B network 145.27.0.0 (not sure where you got classless network) the default mask is 255.255.0.0 or a /16 prefix length. The number of network bits is 16 and the number of host bits is 16. To subnet, we borrow bits from the host portion. If I have 16 bits available what is the largest number of bits I can borrow from 16 to create a new subnet?
Answer: 14
I must leave at least 2 bits for hosts in my new subnet. This results in a /30 prefix length. For CCENT or CCNA you will not use the /31 or /32 prefix lengths. You will learn about these at the CCNP level.
So, coming back to the question, we have the following:
network bits = 16
subnet bits = 14
host bits = 2
# of subnets = (2^14) = 16384
# of valid hosts per subnet = (2^2)  2 = 2
Remember, you can not use the network or broadcast addresses, leaving you only 2 valid hosts in each /30 subnet.
Not sure where they got only 254, but I suspect they only borrowed 8 bits from the host portion, taking the Class B address and making it a /24 or 255.255.255.0 subnet. But this is not the "MAXIMUM" subnets that can be created from a Class B address. The correct answer should be 16,384 subnets.
HTH
Aloha Brian,
Thanks for the reply mate. I worked out the answer just as you did.
But the question comes from a book I am required to use to get my Network + certificate. The question is a review question from the chapter dealing with subneting.
"If you worked on an older network that could not interpret classless addressing, and your network ID was 145.27.0.0, what is the theoretical maximum number of different subnets you could create on this network?"
(I assumed the network is only classful because it can't interpret classless addressing)
Answer (according to the publishers): 254
Since you too confirmed the answer I had, 16382 subnets with 2 hosts (2^14  2) = 16382, I am beginning to think this is either a complete stuffup on the editors part in the book, or we are both wrong.
Kindest regards,
Mike B
16382 subnets with 2 hosts (2^14  2) = 16382,.. a complete stuffup on the editors part in the book ?
NO, in the past, we had to subtract 2 from number of subnetts as well as from number of hosts !
Today, unless you have no ip subnetzero configured on router or you run RIP 1, you do not have to subtract 2 from number of subnetts. this also means you can use subnet zero.
You always subtract 2 from number of Hosts.
Aloha Bouts,
As Martin mention in his post above, you subtract 2 when you have "no ip subnet zero". I believe since IOS 12.0, "ip subnet zero" is enabled by default allowing you to now use the all "zero" and all "ones" subnets. So to clarify,
ip subnet zero
# of subnets = (2^14) = 16384
no ip subnet zero
# of subnets = (2^14)  2 = 16382
either way, the 254 the book has is incorrect. They get 254, by borrowing only 8 bits and assume "no ip subnet zero", so that the # of subnets = (2^8)  2 = 254.
But this is incorrect, in that you can borrow more than 8 host bits to subnet the Class B address 145.27.0.0.
HTH
Hi Brian and Martin,
Yeah, I reakon the book has stuffed up. Found two other questions that have totall wrong answers. I have even contacted the training company that endorses and distributes this book with several examples of my concerns. But as yet have had no reply.
I am currently doing this course through TAFE here in Western Australia. When they come back from our summer break in Feb, I will be contacting my lecturer with my concerns about the quality of this book.
Thanks for your imput guys. You think you must be going mad or just not getting it when no matter how you think about it when your answer is different to a "Professional Training" guide. But when others confirm your suspicions, then you feel a little more confident.
Kindest regards,
Mike B
get W. Odom 's books. they also have a few misprints (get errata from ciscopress) but none related to subnetting.
More, you will get video on subnetting by Odom, and practice test ofr your PC