Sub-sub netting?

Mark Knutson

I have a cisco lab example that appears to require subnetting a subnet due to the intentional assignment of too few bits to accomodate the hosts and subnets in one mask.

Question is, is this sort of thing on the exam? I think I can do a single layer subnetting exercise reliably now, but not the process of taking a subnet and subnetting it...

Maybe I'm just a bit burned out on subnets this week and will be able to figure it out next week after digesting the subnetting process a bit.

Details? Class C address with requirement for 80 hosts in one subnet, 40 in another, and 20 in a third. Total of 5 subnets required. With only 8 bits to play with, a single subnet will not hack it. Or maybe I am missing something, but I suspect not.

dynamik

Those are called variable-length subnet masks (VLSMs), and you'll need to know that for ICND2 or the full-blown CCNA exam. I don't recall that being on ICND1.

kryolla

its the same process whether you have a /24 and you need to subnet that or if you have a /26 and you need to subnet that.

edit: how do you need 5 subnets when there is only 3 requirements and to help you along the subnet with 80 hosts will be a /25 mask so its either going to x.x.x.0 mask 255.255.255.128 or x.x.x.128 mask of 255.255.255.128 so you have 126 hosts to subnet with a requirement of 60 so you can work on the upper half or lower half

Mark Knutson

VLSM sounds right. Its on about the 4th lab into the book, and I was not wanting to deal with it until the second test, but... gonna have to know it eventually, so might as well get to it. I just wanted a few weeks to revel in the glory of knowing the subnetting process from classful addresses.

jason_lunde

There are a bunch of really good explanations on VLSM in the archives within these forums. Also, the trainsignal and cbtnuggets videos do a very good job of explaining them. VLSM is actually pretty straightforward and fun once you get used to it. Good luck with it, and if you have any problems let us know!

Only 7 more dynamik
...then you will implode:)

trackit

if you know how to subnet then you know how to subnet a subnet i dont see how this is any more difficult... If you dont know how to subnet a subnet then you dont know how to subnet

the only thing that you have to remember is that you start with biggest subnet and move to smaller and smaller

networker050184

trackit wrote: »

if you know how to subnet then you know how to subnet a subnet i dont see how this is any more difficult... If you dont know how to subnet a subnet then you dont know how to subnet

the only thing that you have to remember is that you start with biggest subnet and move to smaller and smaller

+1

I've never understood why people have trouble with this.

Mark Knutson

networker050184 wrote: »

+1

I've never understood why people have trouble with this.

Guess I'm a slow learner.

kryolla

Mark Knutson wrote: »

Guess I'm a slow learner.

Dude dont worry about other people just take your time. It doesnt come natural to anybody just keep working on it and you'll get it You can also start on the smaller subnet or the bigger subnet it doesnt matter, I personally start with the smallest.

networker050184

Mark Knutson wrote: »

Guess I'm a slow learner.

Not saying you are slow or anything, I just always see people way over complicating this. If you can already subnet this is the same thing.

Forsaken_GA

networker050184 wrote: »

+1

I've never understood why people have trouble with this.

Because we all make such a big deal out of subnetting. That's why people over think it. It's not until you stop and laugh at how foolish you're being that you really understand subnetting. Once it dawns that it's all just powers of 2, and all of the possible values are quite predictable, it gets easy and you don't need to ask subnetting questions anymore hehe

trackit

kryolla wrote: »

You can also start on the smaller subnet or the bigger subnet it doesnt matter, I personally start with the smallest.

Well.... yes... you can start with smallest subnet, but i will add a little note to topic starter regarding that. I think its more natural (and easyer for a beginner) to start with largest. If you start with largest then you can start your next subnet right where your last chunk left off... but not so when you start with smallest.

For example lets say you have class C subnet 192.168.2.0/24 and you need VLSM it to fit 100, 50 and 20 hosts. You take a first chunk of 128 (192.168.2.0/25) then you can take the next chunk of 64 right after last left off (192.168.2.128/26) and the last chunk of 32 again right after last left off (192.168.2.192/27).

But if you start with smallest you get first chunk of 32 (192.168.2.0/27). Now, you have to know that you cant start your 64 chunk after last left off (192.168.2.32) you have to start it at 192.168.2.64 (192.168.2.64/26) etc...

Sure, this is no biggie for a person how knows his way around subbnetting, but if you are already struggleing with subnetting then i would suggest starting with largest subnet.

billscott92787

I highly recommend starting with the largest subnet. I started with the smallest before and was getting weird ranges. It's not that hard. You are given the main subnet and then told to subnet from there. If you need any further help or explanation let me know. Just keep practicing and you will get it!!

Mark Knutson

kryolla wrote: »

Dude dont worry about other people just take your time. It doesnt come natural to anybody just keep working on it and you'll get it You can also start on the smaller subnet or the bigger subnet it doesnt matter, I personally start with the smallest.

Thanks. I think I understand it, its just tedious. I know how to work the numbers for the most part, but I'm not comfortable until I have an intuitive feel for it. I think I can see how the routers can and or nand the ip address against the mask to get the network or host address. Its just a pain in the butt do do the decimal/binary stuff. Never was good at pencil and paper math.

I guess its good to be able to do, even though its easy to find a computer program that would do it for you in real life--or do we need it for those times we are stuck in a wiring closet with no access to a computer...

It reminds me of learning morse code for my ham radio license. Not necessary an essential skill, one filled with tradition, and it certainly separated the motivated from the casual.

I did a lab that required the use of the subnet ranges I calcuated, and it really helped me see how it all fit together. Now that there is NAT, I imagine most internal networks can just use a 10 network and not have to sweat network/host boundaries that don't line up on octet boundaries, but again, I guess we need to learn things more thoroughly than the average person and take pride in the cert accordingly.

billscott92787

I guess it all depend on your preference. If you plan on taking the exams then I know you have to know it for the CCNA. I'd just practice a good bit with the paper method, then eventually you'll be spitting them out like an english sentence. LOL. You can try subnettingquestions.com - Free Subnetting Questions and Answers Randomly Generated Online (AWESOME site). You can also practice at these two:


Subnetting Quiz #1


Subnetting Quiz #2 (CIDR)


Good luck

colink24

trackit wrote: »

if you know how to subnet then you know how to subnet a subnet i dont see how this is any more difficult... If you dont know how to subnet a subnet then you dont know how to subnet

the only thing that you have to remember is that you start with biggest subnet and move to smaller and smaller

The problems start when you join a company that has setup its 1000 host LAN with a class B /16 mask and you are screwed

trackit

colink24 wrote: »

The problems start when you join a company that has setup its 1000 host LAN with a class B /16 mask and you are screwed

its good to join a company where things are baaad, cause that gives you opportunity to make improvements and earn good reputation

EDIT: i mean, dont you wanna be the guy who replaced the hubs for switches and is concidered a hero for dramatically improving network performance

Mark Knutson

Ok, I finally had a chance to work out the lab and I have it up and running--ping from end to end. There were a total of 4 layers of sub-subnetting required. Thanks for the suggestions.

I think I have subnetting generally mastered, but will need more practice to get it to the speed that is required for testing.

PS. Typically companies don't need to hire you when things are properly set up and running well. The trick is to suss out in the interview just how much trouble they are in, and whether you can get them out of it, as they are understandably reluctant to be completely candid.