up vote 2 down vote favorite
share [g+] share [fb]

I'm trying to get a better understanding of how Token Ring networks operate, but I'm having trouble discerning something.

From what I understand, there are two types of Token Ring networks: wired, and star-wired.
I understand that a "star-wired" configuration requires a MAU that all the clients connect to, but in a "wired" configuration, is a MAU involved anywhere? Or does each computer just connect to it's siblings directly?

I've researched on the internet, but the closest thing I can find are abstract diagrams that don't say for certain whether or not a MAU is involved, and a textbook I have which glosses over the matter entirely.

UPDATE For future viewers: Here's everything you could ever want to know about TR.

link|improve this question
3  
I don't really think there's anywhere you should ask questions about Token Ring -- How hasn't this technology died yet? :) – voretaq7♦ Jan 18 at 21:18
I'm really surprised your school even talks about Token Ring. It should be a quick blurb in the "History" section next to the IP Classes at the most. – Chris S♦ Jan 18 at 21:25
@Chris - It's a technical university, and the class is a Network+ training course. Be that as it may, I still think TR is really cool. Interesting point for those that care - Token Ring's applied throughput of 16 Mbps is actually faster than 3/4 of the Ethernet in the US :) – ajax81 Jan 18 at 22:15
etherealmind.com/myth-fibrechannel-over-token-ring I've not yet figured out if this is taking the piss or not. – Tom O'Connor Jan 18 at 22:28
1  
@ajax81 That's an aggregate network wide throughput of 16Mbps. Ethernet only has network wide limitations when hubs are used. I don't know where you're getting the 3/4 number from, but 10Mb single collision domain networks are getting very rare. 100Mb is on the decline as 1Gb is the only new equipment you can get (in most cases). 10Gb and 40Gb will be common in the next 2 or 3 years. – Chris S♦ Jan 18 at 23:37
show 3 more comments
feedback

2 Answers

up vote 4 down vote accepted

The other answer already says this, but not very clearly.

Token Ring networks always have a single physical ring that the electrical signal travels along, going from one network adapter to the next until it gets back around to the start. There is no (traditional) way to get around this topology.

Originally Token Ring NICs simply had a wire running from one machine to the next, in daisy chain style. If one of the NICs had a problem however it would interrupt the ring and problems ensued.

To get around these single points of failure and the fact that you had to wire from one machine to the next constantly, MAUs were invented. They allow a wire (actually two wires, down and up) to run from the machines to the MAU. The MAU would simply connect one active machine to the next. If a machine was not "active" then the data signals would skip that port and go to the next.

With MAU

(Picture stolen from Wikipedia) You can easily imagine how this would have to be wired without the MAU. The wiring would simply follow the black lines as shown in the drawing.

link|improve this answer
Ah. I see. Thank you very much, Chris. – ajax81 Jan 18 at 22:12
feedback
up vote 3 down vote

In Token Ring, the logical infrastructure view is always a ring/open ring (depending on how you want to look at it).

MAUs just hide it and make a ring inside the device. With a single-wire configuration you only need resistors on the ends of the cable, but when one station disconnects, whole network goes down, unlike with a MAU.

link|improve this answer
Electrically the MAU is a relay (and some other bits) that only opens the ring to let traffic pass through the attached device if one is attached. It isn't strictly necessary, but operating a token-ring network without using MAUs is not really practical. – voretaq7♦ Jan 18 at 21:17
feedback

Your Answer

 
or
required, but never shown
discard

Not the answer you're looking for? Browse other questions tagged or ask your own question.