Previous Table of Contents Next


Hour 10
Token-Ring Basics

“Neither bird, nor plane, nor even frog. It’s just little old me…Underdog.”

The underdog of LAN topologies, Token-Ring, is jeered at by those who don’t use it and venerated by those who do. What’s the straight skinny? For many folks, Token-Ring has given faithful, speedy, and reliable service for many years, but predicting its future is hard. Although Token-Ring vendors have joined the 100Mb ranks by recently announcing HSTR (High-Speed Token-Ring) gear, as well as gig (1000Mb) Token-Ring products in the near future, Token-Ring remains more expensive than Ethernet of similar speed, and it’s getting hard to find technicians who are experienced in deploying and troubleshooting it.


Some folks think that this lack of experienced technicians will be the nail in the coffin for Token-Ring. The argument is that Token-Ring will go the way of the superior yet more expensive BetaMax-type technology: Token-Ring shops will continue to use it for another three to seven years, but new gear, technicians, and parts will become hard to come by, and therefore will be expensive. Eventually, cost justification will sway even the most die-hard Token-Ring devotees to start using other topologies.


One observation is that savvy IT managers are keeping Token-Ring in the mix for now and will distribute it as it breaks—in other words, if it ain’t broke, don’t fix it! Ripping out an entire networking infrastructure all at once can be expensive and dangerous to one’s career if no cost justification exists for it. Nonetheless, IT managers who believe they see the writing on the wall are planning to get rid of Token-Ring as things break or as new networking projects develop.

On the other hand, the availability of High-Speed Token-Ring gear makes it safe and easy to keep doing what they’re doing, only faster. Only time will tell if the availability of High-Speed Token-Ring will trigger its rebirth. As I write this, though, it’s looking pretty grim for the Token-Ring home team.


Regardless of Token-Ring’s future, the fact remains that you’re reading this chapter, so either you’re interested in what all the hoopla is about regarding Token-Ring versus Ethernet or you’ve inherited a Token-Ring network. Token-Ring is well-designed stuff, so be of good cheer. Welcome!

Follow the Rules: Share and Enjoy!

Token-Ring, regardless of its dubious market presence, makes one heck of a great shared network. Unlike Ethernet’s free-for-all rules, where everybody arbitrarily and randomly talks and then retries if they get an error, Token-Ring’s party line has strict rules. Whereas Ethernet is a “be home by 11:00” parent, Token-Ring is the kind of parent who wants to know who you’re going with, where you’re going, who’s driving, and so on. Token-Ring has a lot more rules and therefore has a lot more recovery and manageability.

Just for a moment, think of everybody in a Token-Ring network as being in one big conference room. Suppose there are a lot of people (say, more than 50). How does anybody get a word in edgewise? Think about what you might do in such a situation. You’d probably elect a chair, and this chair might give everyone a turn to speak. This is pretty much what Token-Ring is all about. In large groups, Token-Ring is much more efficient than Creepernet, uh, I mean Ethernet. It’s sort of like using parliamentary procedure rather than family discussions: It’s a wasted effort for small groups, but a lifesaver for large groups.

First, how does Token-Ring work physically? Each workstation that enters the ring can be thought of as filling a seat at the conference table. This is actually how Token-Ring is wired (that is, in a ring rather than in the “star” topology that Ethernet uses). Even though the wiring looks star-like, the path of the wiring actually forms a loop from station to hub to station, as in Figure 10.1. You might be wondering why doesn’t this act the way Christmas lights do? The answer is because all cables come to a hub; if a break occurs, the hub (called a MAU or multiple access unit) can route signals around the break (unless something else is wrong).


Figure 10.1  Even though a hub is present, Token-Ring’s wiring actually forms a ring, which makes it look like a star.

Think of each station that connects to the MAU as sitting down at the conference table (the ring). The first station on the network is called the active monitor (the elected chair), and it acts as a point of contact for others to report errors, maintains the ring’s timing, detects certain kinds of ring errors on its own, and initiates a “do over” if the ring gets certain serious errors. Each station on the network that subsequently joins the ring becomes a standby monitor and can take over if the active monitor leaves the ring.

The active monitor is also responsible for issuing a special packet called the token. Each packet on the wire is handed from Token-Ring card to Token-Ring card, all the way around the ring. Everyone takes turns speaking, one after the other. When a station has the token, it may transmit its own information.


Previous Table of Contents Next