3Com 2000 TR manual Tunneling, Requirements, Token Ring Frames

Tunneling

This section describes tunneling, the method by which Token Ring frames are encapsulated (embed- ded) within Ethernet frames.

Frames are sets of bits that form an elementary block of data to be sent over a communications channel. A frame contains its own control information, including the transmission address and data for error detection. This header information makes certain that the frames arrive with their payload (user data) intact and at the right destination. Simply put, what tunneling does is to take a Token Ring frame and embed it into an Ethernet frame to be sent over a Fast Ethernet net- work.

Requirements

The tunneling scheme must be compatible with Fast Ethernet networks so that a mix of Ethernet, 802.3, and the Ethernet/Token Ring (E/TR) frames may reside on the same segment without interference. It is also important that any end station (server, bridge, router) be able to reassemble the original TR frame. Source routing from Token Ring to Ethernet and back to Token Ring should be supported.

Token Ring Frames

Figure 2-3 shows a simplified drawing of a Token Ring frame.

Tunneling 2-5

	Token Ring Header Information
		Payload
AC FC	nDA nSA RIF (optional)	User Data	FCS
Key
AC–Access Control Field		nSA–Non-canonical Source
FC–Frame Control Field		Address Field
nDA–Non-canonical Destination		RIF–Route Information Field
Address Field		FCS–Frame Check Sequence

Figure 2-3Token Ring Frame.

The Token Ring frame must be specially tagged when preparing it for embedding inside the Ethernet frame. Tagging allows an Ethernet frame to carry:

■User priority information over media types other- wise unable to signal priority information

■Token Ring data in native format (802.5) over non-802.5 media

■VLAN identification information

Tagging requires the addition of a tag header to each frame. This header is inserted immediately following the Destination Address (DA), Source Address (SA), and, if present, the Routing Information Field (RIF). If the source and destination media differ, tagging the frame may involve translation or encapsulation of the frame remainder. Tagging also requires re-computa- tion of the Frame Check Sequence (FCS), which are extra characters added to frames for error control purposes.