I'm getting ready to take 216 and I have a question about subnetting. I've got subnetting down pretty well but there seems to be some confusion about whether you can use all 0's and 1's in the Net ID portion or Host portion. In other words in the real world there are in fact instances where you can use all 1's or 0's but it's my understanding that Micro$oft and Cisco exams stay with "no Net or host portion can contain all 1's or 0's". (no .0 or 255 you know what I mean) I really don't care either way but for testing purposes I need to know what the "official" position is. Thanks as always
Nitro
I love the smell of nitro in the morning

Report this post to a moderator

Freak has a great subnetting guide on his site
www.mcsefreak.com
that should answer your questions. If it doesn't then post the question in his fourms and someone will
hope this helps
For this test you will see 3 - 4 pure subnetting questions.

__________________
Wilbur A+ MCP MCSA-Security MCSE-Security CCNA Security+ ITProject+
-----
IP therefore I am
I think therefore I subnet

Report this post to a moderator

what exactly are you saying? give me an example of an instance .. I think the subnetting is pretty straight forward, maybe the only part of the ms exams that are ..
-neural

Report this post to a moderator

For example say you want to know if a Net ID or default gateway would work on a given subnet. Suppose you're doing a classless address where the 3rd octet is 240 so the first 4 bits 1111 define part of your mask which is /20 so the mask would be for example 255.255.240.0 so far so good. Now your first subnet could be 192.168.16.0 but according to my understanding you can't use (for testing purposes) 0 or 255 in other words 192.168.16.255 wouldn't fly. You know the formula 2^n-2 ? this way you remove the 1st and last addresses which consist of all 1's and all 0's. I hope this clarifies if not I will try and find an actual example. I saw somwhere maybe on a Transcender answer where they said that while it is true that most routers for example won't rout a Net or host portion that contains all 1's or 0's that this isn't always the case. I realize that but all care is what Microsoft "thinks" the right answer is for testing purposes. I'm going to try and find the exact question and get back to you but I think you guys know what I mean. Thanks,
Nitro

Report this post to a moderator

I must be experiencing "pre-traumatic stress syndrome" a serious condition that causes individuals who are studying for MCSE exams to lock themselves up in a room and study until they start seeing things that aren't really there. I went to "Freaks" sight as directed and he says 2^n-2 is the way to go, no 0's or 1's. I couldn't find the Transcender question that I was refering to so maybe I misread their explanation. It's just been bugging me in the back of my mind because I was pretty sure of myself until I saw that. Anyways I don't believe its a big deal because I'm sure that most of the subnet questions won't come down to that however I'm just trying to cover my butt cause 216 is a tough one. Thanks,
Nitro

Report this post to a moderator

Hi nitro,

Relax! Though 216 is the most difficult core for MCSE, don't presure yourself too much. It will only make it worse.

As for your question, from my MCSE Training Kit for 70-216, some guidelines for assigning IP addresses are as follows:
1. The network ID cannot be 127 because it's reserved for loopback and diagnostic functions.
2. The network ID and host ID bits cannot all be "1"s. This is interpreted as broadcast.
3. The network ID and host ID bits cannot all be "0"s. This is interpreted to mean "this network only".
4. The host ID must be unique to the local network ID.
...

There are a few more but they don't apply to what you were asking. With the above guidelines, it's safe to say that Microsoft follows the 2^n-2 formula.

From my experience, one easy subnetting question in the 216 exam would sound something like:

You have a class C IP address of www.xxx.yyy.zzz and you would like to have a x number of subnets with y number of hosts in each subnet. What should the subnet mask be?
Hope this helps. Good luck!

Report this post to a moderator

Thanks Anthonie! thats exactly what I was looking for. Subnetting doesn't bother me it's actually kind of fun. I just wanted to be sure of Microsofts position for testing purposes only and I sure appreciate your answer. That's the neat thing about IT is nobody can possibly know everything so you rely on your peers as well as help others when you know the answer. Thanks again,
Nitro


Nothing exceeds like excess!

Report this post to a moderator

Just be sure to understand the question, is it subnetting or is it supernetting.


Good Luck !!

Report this post to a moderator

Hi all,

I think the question should be precisely like this: no all '1' or all '0' for host portion of an IP address. Because all '0' host address is used for subnet address, and all '1' host address is for sbunet broadcast. For instance, you have a subnet 192.168.16.0/24, the address range is from 192.168.16.0 - 192.168.16.255. But 192.168.16.0 is refered to this subnet and should not be used for any host on this subnet. And 192.168.16.255 is reserved not used by any host either. This is the rule.

On the implementation, Cisco has loosen the rule. That means, you can use all '0' host address on router's interface. This permits more efficient address usage and reduces waste.

Hope that will answer your question

Report this post to a moderator

2^n-2 is the correct for test. 2^n-1 is in practice for Cisco.

Ciao

Report this post to a moderator