Some subnetting queries

[cyberspace]

I understand subetting and can work it out, but I have a few questions.

I am hammering subnetting at the moment for class C addresses so that I can quickly establish subnet masks and ranges without lenghty pen and paper exercises, so saving time in exams.

Anyway, firstly....it occurred to me yesterday, what is actually the benefit of subnetting a class C address (non VLSM manner)? You lose addresses.....does it improve performance?

Also, all addresses with the same subnet mask - can they all talk to each other or is a router required between subnets?

Also, in CCNA the IP-Subnet Zero is typically not used, when assigning a host gateway address, will this be the same for several subnets (eg, 192.168.0.1) with the same mask or will each subnet require a seperate gateway address?

Stuff I should know but have forgotten about and just want to clear up!

Thanks

Systems Administrator
BSc Network Computing, CCNA. Both in training!

 

[qz01gs]

On the CCNA exam you do get questions about subnetting where subnet zero is permitted and also where it isn't permitted.

Cisco dont try and trick you it is clearly stated in the question so practice both methods.

Inb my exam I hardly got any subnetting questions only about 2 or 3.

 

[cyberspace]

Well I use the 2n-2 rule...so when ip subnet zero is permitted I will not need to use this, and just make sure the ranges start at subnet zero instead of subnet 1.


Systems Administrator
BSc Network Computing, CCNA. Both in training!

 

[qz01gs]

good luck for the exam, let us know how you get on.

 

[cyberspace]

Oh I am a while from the exam yet! Just subnetting is the chapter I am working on this week and I really want to drill it in.

Systems Administrator
BSc Network Computing, CCNA. Both in training!

 

[wybnormal]

Subnetting gives you a way to control the use of IP addresses. It's not for performance. Take WAN links for example, why burn 254 address between two routers when you can subnet them to allow just the two iP needed AND provide some nominal security by only allowing those two IPs on the wire. Now you have another 200 address to burn on the rest of the links.

MikeS

http://www.packetattack.com

http://www.packetpress.net

Home of the book "Network Security Using Linux"

 

[KiscoKid]

As previous poster says, subnetting is just a more efficient way of dividing up a bigger subnet. It stems from the day when people just had public addresses assigned to them from the their RIPE LIR and had to make best use of them on their LAN

Of course nowadays a lot of people just use the 10.x.x.x private RFC subnet and you probably could assign huge subnets if you wish to p2p WAN links but I still find that behaviour a bit odd.

Also routers with the same netmask but different subnet ID cannot talk to each other without a router.

 

[cyberspace]

Thanks for that - i get it. Although what I meant was the benefits of subnetting a class C address...for a small company for example. Obviously stuff like access control comes into it etc.

So the firm I work for has 30 PCs, for arguments sake, if i split that into 5 subnets, would it need more routers?

Another thing, in a lab last night, we split the lab into 2 subnets...half the machines were configured on one subnet, half on the other.

Both ethernet interfaces on the rotuer were configured.

So, in some situations where subnets are not necessarily right next to eachother, will one router per subnet be required?

I'm understanding it very well and i'm pleased with my progress, just need to fuklly understand the subnetting routing process...and then move on to VLSM.

Thankyou for posts so far.

Systems Administrator
BSc Network Computing, CCNA. Both in training!

 

[KiscoKid]

A single router could route for 5 subnets if that router

A. Had an appropriate default gateway address for each subnet

B. The router needs to have a local connection (i.e. a layer 2 connection) back to all users in each subnet, i.e. all users should be able to broadcast an ARP to their default gateway to routing to non-connected subnets.

Therefore if the router in question is seperated from these other subnets by, for example, another router then ARP broadcasts won't reach this router and therefore it couldn't service as a default gateway for those subnets.

I hope this makes sense.

 

[cyberspace]

That does make sense, thanks.

Systems Administrator
BSc Network Computing, CCNA. Both in training!

 

[Cluebird]

Rather than one router per interface, you can use a routable interface per network. That is the gist of a Router on a Stick when several VLANs are trunked from a switch to a router. You can also use secondary IP addresses on interfaces and route maps to send traffic out the interface is came in on to a different address range that lives on the same physical interface. There are lots of crazy things you can do to move traffic in and out of the same interface to support multiple networks.