Wendell Odom subnetting (101)

Magic Johnson

His method for calculating the size of the network, subnet and host parts of the address (example below): 8.1.4.5

Find the DDN = /23
Network part = 8 as class A
Subnet part = 23-8 = 15
Host part = 32-23 = 9
Total hosts = 2^9 -2 = 510
Total subnets = 2^15 = 32,768

How ON EARTH am I supposed to calculate all that information, in my head, in under 15 seconds (especially the subnets)?! Are we expected to memorise the 2^ table up to 32? It took me about 30 seconds on paper with a calculator lol.

I know you aren't supposed to mix and match subnetting methods but what am I missing here?

p.s VLSM isn't covered until much later.

networker050184

As they say, practice makes perfect. Once you practice this a million times it comes much easier.

Magic Johnson

networker050184 wrote: »

As they say, practice makes perfect. Once you practice this a million times it comes much easier.

Actually I've just done a 130.x.x.x /24 in about 15 seconds. Proper easy though. Once it goes beyond 2^12 I struggle quite a bit.

bbarrick

You will get to the point where you will see that /23 and automatically know that it your working with 7 bits in the third octet which makes the third number in the mask a 254. I'm hoping that by knowing that 2^8 = 256 that I'll be able to quickly calculate any other powers close to that. 2^9 = 512, 2^10 = 1024 etc...

I understood it by reading Odom's book, became much quicker at it by reading through Lammle's book.

It will get monotonous after awhile and you will see that your dealing with the same method and some of the same numbers over and over again.

DoubleNNs

It's imperative that you understand the math behind subnetting.

However, after a while you'll learn the patterns and start memorizing some of the math - even w/o trying to. Route memorization of the decimal math (not binary) and practice will have you subnetting in your head within seconds. You'll only be writing the problems down as a formality to make sure you don't mess up and miss any easy points on the exam.

draught

While I agree that the math really needs to be understood. I also had problems understanding Odom at first when I was learning. I'd recommending looking into some other subnetting methods as Odom is not the only way.

Magic Johnson

Ok so I was only part of the way through, I have now completed the working with one mask section of subnetting.

I know how to find out how many subnets and hosts on a given IP and mask, and I know how to find the subnet ID/broadcast/range of addresses on a given IP and mask.

Whilst I do think three reasonable sized chapters was a bit contrived giving that in depth base level of how it works certainly made it 'click'. I mean once I got it I got it. I'm practicing now and I can do it in my head quicker than I can writing it down (especially using the magic number technique for finding subnet ID etc)

Bring on the joys of VSLM I guess!

As you guys said, once you see the numbers once they are the same every time. I think memorising is definitely the way to go with this, Odom says it is optional and you can do the math just to be sure but there is much emphasis on the time limit of the exam.

So to summarise: contrived, but gives you a much better understanding of how it all works.

DoubleNNs

I used a combo of memorization of math when I took my CCENT.

Magic Johnson

DoubleNNs wrote: »

I used a combo of memorization of math when I took my CCENT.

I'm just doing some subnetting now, man sucks if you get a small multiplier to work out an interesting octet IP part of 230+ or something.

Would you use multiplication or would you use division in this case (you personally)?

Also, in the exams, are there combination questions for example on troubleshooting? So you get a scenario and it could be anything from the host being on a different subnet (so you'd need to do all these calculations to find IDs and ranges) to configuration (switchport configs, port security etc) to incorrect cabling, to whatever?

DoubleNNs

I've learned how to subnet from so many different sources (Mike Meyers Net+, Todd Lammle, Wendel Odom, CBT Nuggets for MCTS and CCNA) that by now I kind of just.. do it. I'm sure I use my own shortcuts and method by now. So unfortunately, I don't quite understand what you mean (off head) by multipliers on an interesting octet and whether I'd personally use division/multiplication.

However, to answer your 2nd question, the exam will have a lot of "combo" questions. Without giving too much detail, the question will hardly EVER directly ask you to actually subnet. Instead, it'll ask you to troubleshoot something, and in order to answer the question, you'll have to subnet to confirm whether or not the network is set up correctly.
This is why it becomes imperative to subnet quickly. You'll subnet a lot, without that even leading to the actual answer of the question, and then have to continue w/ the rest of the problem afterwards.

Magic Johnson

DoubleNNs wrote: »

I've learned how to subnet from so many different sources (Mike Meyers Net+, Todd Lammle, Wendel Odom, CBT Nuggets for MCTS and CCNA) that by now I kind of just.. do it. I'm sure I use my own shortcuts and method by now. So unfortunately, I don't quite understand what you mean (off head) by multipliers on an interesting octet and whether I'd personally use division/multiplication.

However, to answer your 2nd question, the exam will have a lot of "combo" questions. Without giving too much detail, the question will hardly EVER directly ask you to actually subnet. Instead, it'll ask you to troubleshoot something, and in order to answer the question, you'll have to subnet to confirm whether or not the network is set up correctly.
This is why it becomes imperative to subnet quickly. You'll subnet a lot, without that even leading to the actual answer of the question, and then have to continue w/ the rest of the problem afterwards.

Sorry the magic number method = 256 - the interesting octet. Then say it equals 16 then you have to do multiples of 16 to find the lowest or equal number to whatever number corresponds to the section of IP address (say 234).

Yeah I thought so, but thanks for the clarification!

TechGuru80

Frankly I would know up to 2^10 but you just have to double it. Anything more than that I would imagine is not going to be required. That explanation at least how you typed it is quite confusing.

2^bits = # of networks
2^8-bits = increment (minus 2 = # of hosts)

That is a pretty simple explanation to find the networks and then finding the mask is relatively simple.

Magic Johnson

TechGuru80 wrote: »

Frankly I would know up to 2^10 but you just have to double it. Anything more than that I would imagine is not going to be required. That explanation at least how you typed it is quite confusing.

2^bits = # of networks
2^8-bits = increment (minus 2 = # of hosts)

That is a pretty simple explanation to find the networks and then finding the mask is relatively simple.

It was a direct example from the book. I've memorised up to 2^12 to be safe.

I think I've got it all now, I have no idea what I was on about earlier with division (?!) think I was over-analyzing. It is a lot to take in so far I know once I get my head around finding the information about all the ipv4 stuff I'll have forgotten the previous 2 chapters!

I just need to remember the methods to find the # of subnets then #hosts/subnet and also the magic number method for the subnet ID etc.

Some of the practice questions with the Pearson are worded very sneakily indeed! But yeah you get a scenario or incomplete diagram then 6 potential options, having to either find the # of things or subnet ID/broadcast and ranges. Time is the killer, especially if you don't read the question properly because you want to get it done quickly (argh)