Routing in Lehman's Terms
 

Wondering if you Network Gods can help me understand the basics of routing ips.

I understand that this is a complex world I am inquiring about, but if you could just give me the stripped down version of it, I would really appreciate it.

What are the benefits of a routed network vs WDS?

When do you know, you need to be routed?

What are the disadvantages?

simple stuff like that is what I am looking for.

Hmm. I'll try.

What are the benefits of a routed network vs WDS?

When do you know, you need to be routed?

What are the disadvantages?

I am not for sure if that answers your questions, anyone else wanna correct me, or back any of that up?

First, WDS is NOT routing. It's a bridge. A router passes traffic between two logically separated networks, whereas a bridge passes traffic between two networks which are logically the same. Now, some will try to fool you with WDS routing, but really it's not routing at all, one network is hidden behind a NAT address to ensure correct packet routing.

You know you need to route when: You have two different logical networks (ie, one has the address scheme of 10.0.0.0 and the other 192.168.0.0 or whatever) and you need to connect the two with free-flowing two way communication. You can go the lame-arse NAT route, then if you need to talk with things on the other side of the NAT-ing device you can setup static NAT tables and my god what a PITA.

WDS is good for connecting close proximity offices with the same IP address scheme. That's about it. What are you looking to accomplish?

We also need to know what you're trying to accomplish-

Are you trying to connect to the network as a home user, or are you trying to get your office onto a high-speed internet network? How many users? Public or private IP address? Do you plan to grow?

Those are the basic questions any engineer is going to ask- after that they will handle the technical stuff, network design, which routing protocol to use, security, building the LAN behind your WAN access (if needed), firewalls, etcetera...

There's a lot to it. I don't claim to know a lot about it, I'm a back-bone guy primarily- but I do know a few commands, and can feel my way around a router (Cisco).

If you're just hooking up a few PC's so you guys can network, share files, access the internet, and have some security- that isn't anything that $120 at Best Buy can't fix for you.

Just get a router with a built-in switch & hub, ($79- $99), a few cables, Norton firewall for your PC's, and you're good to go.

Hey D- maybe you can chime in here- Man, I took a Cisco Course, and got a 96 on it, but man- I did not really understand how to do subnet masking... You sound like you know your way around a router- is their a simple way of figuring out the subnet after you know how many hosts/networks you are dealing with and what class of IP address?

Ahhhhh subnet masks... I love it!

Ok, I'll try to break it down for you: We all know an IP address consists of four 8-bit octets (x.x.x.x). Each octet can be represented in binary. For example, 10.11.12.13 is seen by PCs as 00001010.00001011.00001100.00001101 Each octet breaks down like this (going from left to right)

bit 1 = 128
bit 2 = 64
bit 3 = 32
bit 4 = 16
bit 5 = 8
bit 6 = 4
bit 7 = 2
bit 8 = 1


Now that you understand the nitty gritty of an IP address, let's apply the same knowledge to subnet masks.

The different classes of subnets are:

Class A: 255.0.0.0 (1-126.xx.xx.xx 16 million host network)
Class B: 255.255.0.0 (128-191.xx.xx.xx 65536 host network)
Class C: 255.255.255.0 (192-223.xx.xx.xx 256 host network)


A typical 24-bit class C mask looks like this in it's decimal format: 255.255.255.0. In binary it's 11111111.11111111.11111111.00000000. Now that leaves the entire 4th octet for hosts (1-254 since 255 is reserved for broadcast). The first three octets are used to identify the subnet, and the last octet for the host. Now Let's say we have a 27-bit subnet now.... or 255.255.255.224. Binary it's 11111111.11111111.11111111.11100000. That leaves only the last 5-bits open for hosts and the other 27 bits for subnets. So you'll have 30 hosts per subnet and a maximum of 8 subnets.

Get it? It's hard, but if I could actually face to face explain it it would much easier!

^

I actually thought that was pretty good, D.

[EDIT]

Dude, when did you get a bike?
(Vehicle: 1998 Cannondale F700)

Uhhh... in 1998. :smoke:

What happened to the Feugo

I figured I'd rotate out my car collection. You'll see it change from time to time. Not often, but every now and again. I really love my Canondale... best bike ever!

You Rock man.

:lol:

And once you learn how to subnet correctly IPv6 will become the law of the land and we'll all be mastering hex.

Ain't IT great!

Yes, but hard to get into. There is a ton of information for newbie's (like myself) to learn...that's the frustrating part.

Yeah...

That's about as simple as getting one's heel in their mouth..

They showed us a "cheat" that I thought I understood- well, I did understand it well enough for that portion of the test not to wax my grade- but, I didn't REALLY understand the principle behind why I was doing it the way they said do it..

Know what I'm saying?

IP v6! HA! They've been threatning us with that for nearly 10 years now! Thanks to our dear friend NAT, v6 is still a long way off me thinks. Besides, Hex is fun, remember the good 'ol IPX network names?!

CIA - so do you get the principal behind it now?

I remember my first pass at subnetting...I just said "yea right". Then one day it just clicks.