Ethernet frame confusion

Bob_the_Goon

Hi,

I'm currently reading through the ICND1 Official Exam Cert Guide (Wendell Odom) at the mo and have got up to the Ethernet Framing section and wandered into a little bit of confusion over the Type Field.

The book states that when a station receives an ethernet frame and starts to examine it, it checks the type field in order to see what packet is lurking inside of the data field. This concept I understand, however, the book says the device can recognise if the frame is encapsulating a IP packet, a OSI packet etc etc.

My question is this. What is the difference between an OSI packet and an IP packet? Is there one? The book seems to suggest that this is the case.

I personally would have thought that all packets IP packets are OSI packets as Internet Protocol is a part of the OSI model. Am I missing something fundamental here or have I just misinterpretated the book.

Cheers in advance!

dtlokee

The type field contains a code that specifies what L3 protocol is inside the frame, in theis case the use of "OSI Packet" is a bit confusing because you are thinking about the OSI model, when in fact there is a "OSI Protocol". If the frame contains an IP packet it will have type code 0x800.

http://www.cisco.com/en/US/docs/ios/12_2/ibm/vol1/command/reference/br1fethc.html

This contains the type codes for some different protocols.

freetech

OSI, in modern usage, refers to the concept of handling network packets in a uniform way. When it looks inside the packet and sees a packet of a certain type, it recognizes that differences exist in how the packet is handled, depending on the protocol it belongs to.
Don't get too hung up over the OSI framework and the differences in protocols. The CCNA/802 exam only talks about TCP/IP. Don't worry about anything else for the exam.
My advice is, mark the section in the book that "you can't get your head around", and skip it. Believe it or not, as Odom goes on, you'll begin to see why he is making some of those distinctions in that section.
IMHO, Odom gets kind of stuck in the weeds in that section. It really bogged me down. Once I understood a little more about the routing concept, that part began to make sense.
So go around it, and come back to it after a few chapters and it may make sense.

Cucumber

I think Wendell Oddom is talking about the set of OSI protocols like ES-ES, IS-IS, etc.
You will read a bit more about them when studying for the BSCI exam (in the CCNP track).

When an ethernet frame carries any of these protocols (ES-ES, IS-IS,etc), the ethernet frame contains a protocol type of 0xFEFE (or is it 0xEFEF) I cant remember. Look it up, dont trust me on this kind of details.

Why are they called OSI protocols?, for what I remember, they are the protocols that the OSI itself created. You know TCP-IP was created by the military, IPX-SPX by Novell, and ES-ES and IS-IS were created by the OSI.

Please correct me anyone if Im wrong.

Best regards

Bob_the_Goon

Smashing. Thanks for the info gentlemen! Much appreciated.