CCNA study guides and books

madhacker

Hi everyone. I am starting my training for CCNA certification tomorrow. The book they will be following is Todd Lamle CCNA Routing and Switching Study Guide. I already got the book and read about 230 pages till now. I'm on the IOS chapter presently. I understood about half of it. I got the subnetting and the TCP/IP part but am confused a bit about collision and broadcast domains but I will get it. I am more comfortable with the detail oriented parts like how encapsulation and de-encapsulation takes place and calculating subnets and valid host addresses in subnets etc.

My question is if there are additional books for this which may be useful but are not essential. I want to build a good foundation since I tend to understand in relation to numbers or logical steps. Is there any book which would make the various parts seem more logical in that I understand the underlying logic. The book by Todd Lamle is extremely interesting and I could probably get by with it but are there other books that can be read out of interest?

TWX

Collision - think of old-fashioned radio. In radio, with humans using radios, only one party can transmit at a time because only one person can be listened-to at a time. It's also like a conversation between people standing in a circle, only one person can be listened-to at a time, and when more than one person speaks at a time the rest of the group has trouble listening to both conversations.

CSMA/CD (Carrier-Sense Multiple-Access Collision Detection if I am remembering correctly) is used on segment of a network where multiple parties could each want to use the wire at any given time. Go back to the days when there was not an individual cable from a switch to each computer, when Ethernet used coaxial cable to link computer to computer to computer. In those cases, if someone wanted to 'speak' on the network they had to first check if anyone else was speaking on the network, and if not, they would initiate a preamble ("I am going to speak now!") before initiating the header. If they detected usage on the line they would wait. If they started a preamble but then detected someone else's preamble, they (and presumably the other party) would back-off and wait a random interval (so they wouldn't both have the same timeout) to try again, where one would start, then the other would detect the transmission, then when it was done, do its own transmission. This system where there could be multiple parties (ie, more than the two ends of a modern cabling between a switch and a PC) is called a collision domain, and when there are more than two parties in a collision domain they must use half duplex to either speak, or listen (ie, transmit or receive) but not both at the same time.

Switches are effectively special-purpose computers that limit the number of devices in a collision domain to two, the PC, and the one port on the switch that the PC is plugged into. This means, that since computers are much better at listening and speaking at the same time, that a computer can both transmit 100% of the time and can receive 100% of the time, since the only other party that will ever transmit on the cable besides them is the one at the other end of the twisted-pair cable. This is made possible by the fact that a PC always transmits on one set of conductors in the cable, and the switch always transmits on the other set of conductors in the cable.

Broadcast - now, like TCP/IP as a Layer 3 concept, Ethernet as a Layer 2 concept has a broadcast address, and packets sent to the broadcast address have to at least be listened-to by all stations, even if the station does not act on them. Additionally since the amount of time that a PC or a switch retains knowledge of the other devices on the network is limited, information that is intended for specific destination might initially be flooded to all devices until the recipient device acknowledges that it is the recipient, at which time the switches along the path will firm-up their knowledge of how to reach a given destination, so that they forward packets through a specific path instead of flood them everywhere. The section of the network in which flooding may occur is called a broadcast domain. Switches only learn the MAC addresses of devices by listening to those devices when they transmit, so until a destination transmits (and after the record ages-out of the MAC table) the network does not know how to reach the device, hence the need to flood frequently. This is true regardless of what Layer 3 protocol is used- happens with old Novell IPX, happened with Microsoft's NetBEUI, happened with Apple's Appletalk, etc.

TWX

Oh, as far as books go, I really am liking Scott Empson's CCNA Routing and Switching: Portable Command Guide Third Edition. ISBN 978-1-58720-430-2. It's essentially a breakdown of commands without the explanation, makes for a good command reference when you're attempting to do the exercises after having read the curriculum without having to dig through a larger textbook.