Here is a question that was asked in my networking class...

When would it be appropriate to use Token Ring? Give examples.

What I want to answer:

Honestly, with switched ethernet, I do not think that it would be appropriate to use Token Ring.

Places still use that?

I once lost a Token in the ethernet when the Ring broke.

Quoted: Here is a question that was asked in my networking class... When would it be appropriate to use Token Ring? Give examples. What I want to answer: Honestly, with switched ethernet, I do not think that it would be appropriate to use Token Ring.

the only time you'd ever want to use token ring these days is for legacy mainframe equipment that doesn't have an ethernet interface. Otherwise it's just junk compared to switched ethernet.

Quoted: I once lost a Token in the ethernet when the Ring broke.

According to the book, ten meg ethernet is good and cheap...

While... 100 meg ethernet is expensive...

The book is at leat five years, if not ten years outdated on a lot of subjects...

Quoted: We use token ring networks for the modules in our microlithography workcell clusters at work.

Is that a legacy network?

How old is the equipment?

The answer that they're looking for is for applications where network nodes have to be able to communicate on regular basis with no tolerance for varying degrees of latency as a result of network traffic.  An example might be, say, hospital monitoring equipment.  The reason is that in Ethernet large traffic volume can lead to problems when collisions come into play and latency can rise sharply once the Ethernet segment becomes saturated.

Switched Ethernet does indeed more or less eliminate this problem, IF each node is plugged into a switched port.  If a number of nodes share a switched port by way of a hub, obviously collisions come back into play.

Of course, switched Ethernet can bog down with heavy traffic too - but that's dependent on the capabilities of the switch itself to accept and switch frames, rather than on collision handling.  As such, switched Ethernet can handle a lot more, *but given enough traffic can still be brought to its knees*.  Token Ring still maintains an advantage in that it maintains a very regular and very predictable amount of network latency regardless of the traffic on the network.

So, in short, you're right, but wrong.  Token ring is still superior to any manner of Ethernet in terms of "regularity"; however, in practical terms, you are unlikely to have any problems with switched Ethernet if you use the appropriate switches and network configuration for the job.



ETA:  I forgot about FDDI, which is a type of token ring, but not IEEE 802.5 / IBM Token Ring.  That can be good for backbone applications.

ETA2: Primarily because a) it's fiber and b) it's redundant and self-healing, which 802.5 is not.

Quoted: The answer that they're looking for is for applications where network nodes have to be able to communicate on regular basis with no tolerance for varying degrees of latency as a result of network traffic.  An example might be, say, hospital monitoring equipment.  The reason is that in Ethernet large traffic volume can lead to problems when collisions come into play and latency can rise sharply once the Ethernet segment becomes saturated. Switched Ethernet does indeed more or less eliminate this problem, IF each node is plugged into a switched port.  If a number of nodes share a switched port by way of a hub, obviously collisions come back into play. Of course, switched Ethernet can bog down with heavy traffic too - but that's dependent on the capabilities of the switch itself to accept and switch frames, rather than on collision handling.  As such, switched Ethernet can handle a lot more, *but given enough traffic can still be brought to its knees*.  Token Ring still maintains an advantage in that it maintains a very regular and very predictable amount of network latency regardless of the traffic on the network. So, in short, you're right, but wrong.  Token ring is still superior to any manner of Ethernet in terms of "regularity"; however, in practical terms, you are unlikely to have any problems with switched Ethernet if you use the appropriate switches and network configuration for the job.

Yes,

Token ring is superior to shared ethernet. Switched ethernet does eliminate the shortcomings of shared ethernet, and yes, it can become bogged down.

However, with the processing power of the switches on the market today and the advent of Gigabit ethernet to the desktop and ten gig to the core.... do we really have to worry about a properly configured network becoming bogged down?

The most common implementations of token ring have speeds of 4 and 16 Megabits a second.
I know there are standards for 100 and 1 gig implementations, but is equipment even manufactured for this?

Personally, I dont see myself ever recommend token ring.

We are putting 2 48port 100gb switches in one of our computer rooms.  
And all T1 are being replaced with fiber.

Quoted: According to the book, ten meg ethernet is good and cheap... While... 100 meg ethernet is expensive... The book is at leat five years, if not ten years outdated on a lot of subjects...

LoL.   We had token ring up until about 2 years ago.  4mbit token ring, no less.  

.

Quoted: LoL.   We had token ring up until about 2 years ago.  4mbit token ring, no less.

I feel so sorry for you...

At least you did not have 10 base2, coax...that shit is a lot worse!

Quoted: Quoted: LoL.   We had token ring up until about 2 years ago.  4mbit token ring, no less.   I feel so sorry for you... At least you did not have 10 base2, coax...that shit is a lot worse!

Yeah, thank gawd I've been offsite for 10 years and didn't have to put up with it.


Not only that, but we had FIBER token ring, so we spent the BIG BUCKS for our "blazing" 4 mb ring.

(as I recall, the ring could do 16, but we were limited by some cards that only went 4)

Even FDDi at 100 mbit/sec doesn't stack up against ethernet. Gigabit ethernet has hit commodity status, too. I think L2 gigabit switches are going for about $20/port.

Quoted: However, with the processing power of the switches on the market today and the advent of Gigabit ethernet to the desktop and ten gig to the core.... do we really have to worry about a properly configured network becoming bogged down?

In a word - yes.  If x nodes hammering your switch will slow it down below the threshold you have for network reliability, then its only a matter of time until you're dealing with x + 1 nodes.  Then you buy a new switch, and increase x.  Which, BTW, is perfectly fine - if x won't ever be reached, which is the case with many networks.  Large networks may be able to cause problems, which is why Cisco has such a wonderful range of products from the small to the nuclear-powered models.

As far as speed is concerned, keep in mind that Token Ring excels with applications that use lots of relatively small messages that need to be delivered on a regular basis.   In such applications, the actual bandwidth of the connection may play second-fiddle to the overhead of CSMA/CD and frame switching and processing by the switch.  When larger frames are sent like in FTP transfers that overhead is a smaller part of the overall transmission time, and Token Ring's low bandwidth will come into play more.

Without going too much farther into this, 802.5 is a very specialized but IMO NOT obsolete technology.  In some very specific situations (high-reliability small-message networks) you really can not beat it.  Of course, I'm one of those weird IT people that believes that most technologies available can be useful when properly applied, and I'm not very quick to dismiss technologies as obsolete or worthless.

I find more and more that I am in the minority.

Quoted: Even FDDi at 100 mbit/sec doesn't stack up against ethernet. Gigabit ethernet has hit commodity status, too. I think L2 gigabit switches are going for about $20/port.

Depends on if you're going to lay redundant Gigabit paths and if your equipment will fail-over automatically (I hope so for a backbone).  Also, I'm not sure how well Gigabit works over longer distances, between buildings, across the street, etc...  places where fiber can excel.

ETA: For some reason (tired) I assumed you were talking about copper gigabit. If you're talking about fiber, disregard the bit about fiber, but the redundancy and failover need to be addressed as well.

Lots of factors, depending on the needs of the network and the users.  I don't like "one size fits all" approaches, can you tell?

Quoted: Quoted: However, with the processing power of the switches on the market today and the advent of Gigabit ethernet to the desktop and ten gig to the core.... do we really have to worry about a properly configured network becoming bogged down? In a word - yes.  If x nodes hammering your switch will slow it down below the threshold you have for network reliability, then its only a matter of time until you're dealing with x + 1 nodes.  Then you buy a new switch, and increase x.  Which, BTW, is perfectly fine - if x won't ever be reached, which is the case with many networks.  Large networks may be able to cause problems, which is why Cisco has such a wonderful range of products from the small to the nuclear-powered models. As far as speed is concerned, keep in mind that Token Ring excels with applications that use lots of relatively small messages that need to be delivered on a regular basis.   In such applications, the actual bandwidth of the connection may play second-fiddle to the overhead of CSMA/CD and frame switching and processing by the switch.  When larger frames are sent like in FTP transfers that overhead is a smaller part of the overall transmission time, and Token Ring's low bandwidth will come into play more. Without going too much farther into this, 802.5 is a very specialized but IMO NOT obsolete technology.  In some very specific situations (high-reliability small-message networks) you really can not beat it.  Of course, I'm one of those weird IT people that believes that most technologies available can be useful when properly applied, and I'm not very quick to dismiss technologies as obsolete or worthless. I find more and more that I am in the minority.

We can both agree that is not suitable in this day and age for a network of desktop computers?
IE employees surfing the web....e-mail...



It may be suitable for networks that involve highly specialized equipment, but is the technology commonly available and being developed and supported by vendors?

I know that I can buy interfaces for a Cisco router that will do token ring, so I can link an ethernet  network to a token ring network.... But that is all that I am familiar with.

Quoted: We can both agree that is not suitable in this day and age for a network of desktop computers? IE employees surfing the web....e-mail...

Absolutely.  For those applications I would suggest 10Mb Ethernet or 100Mb FE like anyone else would.  I'd suggest switched Ethernet for security reasons rather than performance reasons.  Basically, the home network scaled up a bit is all.

In fact, inside a building, I wouldn't deviate very far at all from that school of thought unless you had a specific reason for it.

Here is what I posted

When would it be appropriate to use Token Ring? Give examples...

In the old days of shared ethernet, and even in the days of early switched ethernet, I could see where it might be appropriate. Token ring is supperior to shared ethernet in many respects. But those days are long gone.

In this day and age, ethernet switches are cheaper, faster, more robust, and more reliable than ever. A switched ethernet network, if designed and implemented properly, can be scaled up to thousands of users very easily.

Is new token ring equipment even on the market? Is it cost effective, is it supportable? Workstations do not come with built in token ring NIC cards, they come with built in ethernet cards. Did high speed token ring even make a dent in the market?

From everything I have read, I just do not think that it would be in the best interest of most organizations to consider this technology.

Here is a good article from 1998, describing why organizations began migrating away from this technology:

http://www.bcr.com/bcrmag/1998/03/p36.php

Cisco makes a very good case for staying away from Token ring:

http://www.cisco.com/en/US/netsol/ns340/ns394/ns74/ns149/net_value_proposition09186a00800c92b9.html



This reminds me of ATM/LANE. ATM is a very good technology for a network backbone. One does not have the spanning tree issues that they have with ethernet. The problem is adapting it to an ethernet world through LANE emulation. This is where it gets horribly complex. I should know, as I had to work with it for a number of years. I understood it and I was quite good with it, but it was not an easy technology to master. Ethernet is 100 times easier to learn, and eaier to deal with!

Gigabit ethernet took the wind out of ATM/LANE's sails. Cisco stopped developing the technology.

Sure, technologies like token ring and ATM/LANE are good technologies when implemented properly. But we must take cost, training, support, and commercial availability into account.

Quoted: Quoted: Even FDDi at 100 mbit/sec doesn't stack up against ethernet. Gigabit ethernet has hit commodity status, too. I think L2 gigabit switches are going for about $20/port. Depends on if you're going to lay redundant Gigabit paths and if your equipment will fail-over automatically (I hope so for a backbone).  Also, I'm not sure how well Gigabit works over longer distances, between buildings, across the street, etc...  places where fiber can excel. ETA: For some reason (tired) I assumed you were talking about copper gigabit. If you're talking about fiber, disregard the bit about fiber, but the redundancy and failover need to be addressed as well. Lots of factors, depending on the needs of the network and the users.  I don't like "one size fits all" approaches, can you tell?

Well...

You could use spanning tree... it does work, when used properly!

Or...why not use layer three to manage your redunant links? Switches such as the Cisco 6500 series with the supervisor 720 are more than adequate to route the traffic on the core of most networks.

You could route to the edge switch closets in your network as well... That is if some of your edge switches can do layer 3.

As far as distance limitations...gig or 10 gig over fiber makes disatnce limitations for most networks a non issue. For the real long haul circuits, I belive you can do IP over SONET.
(I am not very familiar with a lot of long haul technologies.)

We had token ring until about 4-5 years ago.

I remember the company I worked for back in the day switched from Token Ring to shared Ethernet in 1987.

This reminds me of SATA/SCSI arguments. SATA is cool, but the intelligence of the SCSI protocol is what makes SCSI better, IMO.

Wasn't it destroyed in the fire of Mount Doom at Mordor?

Bilster

The only reason that it is appropriate for some companies to use is the cost of converting over to an ethernet setup.  For the most part a token ring setup is bullet proof and as long as it continues to work there is not much reason to switch it out.

Quoted: Quoted: At least you did not have 10 base2, coax...that shit is a lot worse! That was the first home network I had.  We had 5-6 machines in our house running a BBS, all using coax and LANtastic.

Oh yeahg baby!!!

Bring on the Wildcat cards!!

Quoted: Quoted: Quoted: At least you did not have 10 base2, coax...that shit is a lot worse! That was the first home network I had.  We had 5-6 machines in our house running a BBS, all using coax and LANtastic. Oh yeahg baby!!! Bring on the Wildcat cards!!

WWIV rules (ruled?)!

Quoted: Quoted: At least you did not have 10 base2, coax...that shit is a lot worse! That was the first home network I had.  We had 5-6 machines in our house running a BBS, all using coax and LANtastic.

LANtastic... wooo boy

Back when the world was young.

You forget just how big a pain in the ass this stuff was a few years ago.

In my experience, the inherent problems associated with configuring and especially troubleshooting token ring outweigh any theoretical reliability increase compared to Ethernet.

I managed several TR networks years ago. The adapters were always failing, and about weekly we'd have a serious problem where some device would reboot and set its card to 4 mbps and slow the entire network to a crawl.

Later I did some consulting for an industrial plant that had a huge TR network; they were constantly running down problems and tracing cables in that data center.  When they finally converted to a switched Ethernet they were amazed at how much more reliable it was

I don't doubt there still may be a place for broken ring in the IT world, but I haven't struck it yet.

I think MCA/Universal Studios in Universal City(aka Universal Studios, there actual corporate offices), a suburb of 10 miles west of Los Angeles has an extremely complicated topology of ethernet and token ring.  According to the network guy, each of these networks can talk to each other. Sorry can't elaborate any more because I'm not a network guy, and I only spent 5 minutes of causal banter with this guy.

I thought alot of bank used it becuase it is impossible to corrupt data.

Quoted: I thought alot of bank used it becuase it is impossible to corrupt data.

Nope, I betcha that it is more of an economics issue.  By switching over to ethernet they would have to rewire the entire place.  That costs quite a bit of money cause not only do you need to switch out the wiring, but you would need an ethernet switch and you would most likely need a replacement network router unless you have one of the ciscos that are setup to talk to both ethernet and token ring.  

IBM, which invented and promoted Token Ring, switched to ethernet several years ago for all internal  uses. TR is still used in legacy applications that cannot tolerate ethernet, and by customers who do not have a pressing economic need to change, or who have a situation that cannot tolerate the high number of packet collisions common with ethernet, as previously stated.

Modern switched ethernet is collision free when properly configured...
Also, there is a night and day difference from the equipment that was sold in the late'90s.

I think Cisco makes a very good case to switch on their website...

A popular misconception of some proponents for maintaining a Token Ring infrastructure is the belief that sticking with Token Ring is a "zero-investment" decision. In reality, the decision to remain with Token Ring implies a continued investment in the technology, with the purchase of new network interface cards (NICs) for each new PC installed, in addition to the purchase of new switches and routers to support the demand for increased network bandwidth. In most technology migrations of this type, financial metrics usually dictate a migration over time. However, for Token Ring users who are not already well into a migration away from that technology, time may well be short. The Token Ring migration is often coupled with other infrastructure changes, such as the elimination of legacy protocols in favor of IP or the refresh of desktop PCs.

One common stumbling block cited about migration to Ethernet is the existence of shielded twisted-pair (STP) wiring. STP was the recommended wiring for Token Ring networks, and many enterprises have massive investments in STP. Many network managers believe that a migration to Ethernet, which typically utilizes Category 5 unshielded twisted-pair (UTP) wiring, will require a complete rewiring of all buildings. This is not true—the Ethernet standard explicitly supports STP wiring, and inexpensive solutions exist for converting the connections. As a result, organizations with STP cabling can migrate to Ethernet without any rewiring in most cases. Both Cisco Systems and IBM have certified the operation of a selected number of their Ethernet products over STP.

Cisco has already announced the "end of sale" of many Token Ring platforms, and many other Cisco Token Ring products are rapidly approaching their end of life because of a greatly diminished number of customers still using Token Ring products. This lack of demand, in turn, results in an increasing difficulty in obtaining the necessary components for these products. As the popularity of the technology fades, this trend is inevitable and the Cisco experience is, no doubt, also occurring at all the remaining Token Ring vendors as well.

The message from industry analysts and organizations that have undergone the migration is clear—make the decision to migrate right now. To delay the decision means risking your ability to react in the future to deploy new networking applications and new networking technologies or even to maintain your current Token Ring network. In the long run, delaying the migration will cost more and will place your network at higher risk than if you get started today.

http://www.cisco.com/en/US/netsol/ns340/ns394/ns74/ns149/net_value_proposition09186a00800c92b9.html

How about the Tolkien ring?


Preciousssss, myyyy precccciousssss!!!