3 of 3
4/25/07
#15389
724-746-5500
blackbox.com
About Token Ring.
Token Ring was developed in the early 1980s by IBM
®
. In
1985, the IEEE released document number 802.5, which formally
accepted Token Ring as a standard. Although its popularity has
been eclipsed by Ethernet, Token Ring is nonetheless a robust
networking standard and is still operating reliably in many
legacy networks.
Token Ring is most often used in IBM mainframe environ-
ments. IBM provides Token Ring connections on much of its
mainframe computer hardware, and its software enables PCs
and mainframes to act as peers on the same network.
The name “Token Ring“ is somewhat misleading. Although
its name implies a physical ring shape, Token Ring is actually a
star-based topology. Logically, the network operates as a ring—
“tokens“ bearing information are passed from node to node
until the token travels completely around the ring.
Token Ring networks are interconnected via passive and
active hubs called Multistation Access Units (MAUs). MAUs are
the fail-safe components of Token Ring networks. If a cable or
a connection goes bad, the MAU will drop that node from the
ring. This fail-safe hub arrangement prevents a break from
taking down the entire network. Ring-in/ring-out ports in each
MAU can be used to connect a ring to other rings.
Token Ring repeaters are used to extend distances within a
ring. Repeaters can increase either main-ring or lobe lengths in
a Token Ring LAN. The main-ring length is the distance between
MAUs. The lobe length is the distance from MAU to workstation.
Token Ring cabling.
The original Token Ring specifications called for shielded
twisted-pair (STP) cable using either a DB9 connector or a unique
square connector called the IBM data connector. Later, Token
Ring was adapted to use conventional unshielded twisted-pair
(UTP) cable with RJ-45 connectors. The most common kinds of
Token Ring cabling in use to day are Type 1 and Type 6 STP as
well as Type 3 UTP.
Type 1 shielded twisted-pair (STP) cable is the original wiring
for Token Ring. In Type 1 cabling, each wire is constructed of 22
AWG solid copper. Type 1 cable is not as flexible as Type 6 cable
and is generally used for long runs in areas where twists and
turns are less likely, such as in walls or conduits.
Type 6 Token Ring cable is a lighter, more pliable version of
Type 1 cable. It’s constructed of two stranded 26 AWG copper
pairs that are surrounded by an overall braided shield. Type 6
cable is commonly used in offices and open areas, and its flexible
construction enables it to negotiate multiple twists and turns.
Type 3 or UTP Token Ring cabling uses the same twisted-pair
CAT3, CAT5, or CAT5e cabling with RJ-45 connectors as 10BASE-T
Ethernet.
Technically Speaking
Attaching older Type 1 Token Ring to UTP Token Ring
requires a balun or adapter.
Two speeds.
Token Ring operates at either 4 or 16 Mbps. While you can
mix speeds within one network, each individual ring can operate
at just one speed. Simply put, you cannot have 4- and 16-Mbps
devices operating in one ring.
Migrating from Token Ring to Ethernet.
Although it’s proven to be reliable and fault-tolerant, Token
Ring has largely been supplanted by Ethernet, and Token Ring
network devices are no longer widely available. Additionally,
Ethernet LANs are less expensive to build and operate than
Token Ring, Ethernet can provide far greater bandwidth, and
many networking products simply do not support Token Ring
at all. The reality is that Token Ring is dying. Even IBM has
walked away from this standard.
For these reasons, system administrators usually choose to
migrate their legacy Token Ring networks to Ethernet by adding
new network segments as Ethernet rather than Token Ring.
There are three different ways to move from Token Ring
to Ethernet by interconnecting the networks. The method you
choose depends on whether you’re connecting a LAN or a WAN
and how much you want to spend.
Keep in mind that no matter how you connect the two net-
works, they both must use the same protocol. In other words,
they must speak the same language, such as IP or IPX
™
.
An economical way to combine Novell
®
IPX based Token Ring
and Ethernet networks is to have them share a server. Just install
either a Token Ring NIC in the Ethernet server or an Ethernet
NIC in the Token Ring server. Then change the server’s config-
uration to enable the internal router program. While this
method is low cost, it adds traffic and overhead to the server,
which can decrease overall network performance.
An easy, standalone hardware solution is to simply use an
Ethernet to Token Ring bridge such as a BLACK BOX
®
Ethernet
to Token Ring Adapter (LBU9001-US, LBU9002-US). Just connect
the bridge to your Token Ring MAU and your Ethernet hub and
make sure that both networks have the same network
identification—in essence, they will become one logical
network. Also, since this is a solution for LANs, both networks
must be close to each other.
Another choice for connecting Token Ring and Ethernet is to
use a router to connect the two networks over a wide-area link.
The router enables each network to maintain its own unique
network identifier.