08 - Network Architectures

NETWORK ARCHITECTURES (CONTINUED) PREPARED BY ARVIND SHARMA Token Ring Token ring was developed in the 1970s by IBM and uses a token-passing architecture. It follows IEEE 802.5 standard and uses a star physically topology. However, token ring uses a logical ring to pass the token (a small frame) from station to station. Each computer or device must be attached to a central device or concentrator called a multi-station access unit (MAU or MSAU). These MAUs or MSAUs are similar to hubs and also add fault tolerance to the network by using relay switches (also called bypass switches). If a single device fails, relay switch in MSAU will bypass that failed device without interrupting rest of the token-ring network. Token Ring uses a formula called Beaconing that detects and helps to repair these network faults. When a device detects a problem in data transmission (e.g. broken cable), it sends a beacon frame to notify the problem. Data tokens will not be transmitted until the faulty device or cable is detected and ring connection is restored. Beaconing initiates a process called self-correction, where notified devices automatically perform diagnostics. Finally, MSAUs through electrical reconfiguration try to restore network around the failed device or devices. Data frames on a token ring network are transmitted from device to device along a logical ring as shown in figure 8.1. Only one device that got a token can transmit data at a time in either a clockwise or anti-clockwise (most commonly used) direction. Every device connected to a logical ring will check the frames intended destination. If frame is intended for a device, it changes 2-bits in the frame to indicate that data has been read. Otherwise it will send to the next device, until received by the destination device. Finally, the frame will travel along the ring to the source device. It will verify that data has been read by the destination device and will remove data from token (frame). If device wants to send more data, it will attach new data to token and send to the destination. Otherwise, it will release the token to move freely along the ring. This token will keep on moving and can be picked up by a device that wants to transmit data. Any device that receives the token from the preceding device, call it as nearest active upstream neighbour (NAUN). It passes token to the next device on the ring that is called its nearest active downstream neighbour (NADN). Figure 8.1: A View of Token ring To better understand the Token ring frame transmission can be summarised as below: 1. Token will move freely in anti-clockwise or clockwise direction 2. Sending device will wait for the token Internetworking & Middleware 1 -Network Architectures (continued) 1/4 3. Sending device will receive token and attaches data to it 4. It will transmit frames (token and data) 5. Frame will move from device to device until received by destination device 6. Destination device will copy the transmitted frame and modifies two bits (sets Address Recognized and Frame Copied indicator bits) 7. It will send the modified frame back to sending device 8. Sending device will examine the modified frame to verify that data is received and acknowledged 9. Finally it will remove data from the modified frame and releases token to move freely along the ring 10. Other devices can pick the token, if they want to send data In token ring networks, certain high-priority devices can send data more frequently than others using a priority system. Token ring frames carry two fields called the priority field and the reservation field that control priority. Only devices with a priority equal or higher than the priority value of a token can take it to transmit data. If a token is in use, only devices with a priority value higher than that of the sending device can reserve the token for the next turn. Devices that modify a token's priority field must restore previous priority as soon as transmission has been finished. Token ring network interface cards are different from Ethernet network cards and can run at 4Mbps or 16Mbps. In new version of token ring also called fast token ring, data can be transmitted at 100 Mbps. Token ring network devices are also called lobes and often use STP and UTP cables to directly connect to a MSAU. Many MSAUs can be further connected together to form one large ring as shown in figure 8.2 Figure 8.2: A typical token ring network consists of devices connected to MAUs in a physical star configuration Live animation and more information about Token ring can be found at the following web sites: http://www.datacottage.com/nch/troperation.htm http://www.netbook.cs.purdue.edu/anmtions/anim06_3.htm http://www.netbook.cs.purdue.edu/anmtions/anim06_4.htm http://www.ii.metu.edu.tr/~ion504/demo/lan/html/medium.html Fibre Distributed Data Interface (FDDI) The Fibre Distributed Data Interface (FDDI) standard was developed by ANSI in 1980s to allow transmission of digital data over the fibre-optic cable. Recent IEEE 802.3ae specifications allowed FDDI Internetworking & Middleware 1 -Network Architectures (continued) 2/4 to transmit data at a speed of 10 Gbps (10,000 million bits per second) up to a distance of 40 Kilometres. However, at a transmission rate of 100Mbps, it can transmit data up to 100 Km. There were more and more applications everyday that demanded high bandwidth and transmission of data to longer distances. Ethernet and token ring technologies were not able to satisfy these demands. FDDI provided a reliable and high-speed solution for real time applications used in heavy traffic networks. Like token ring, FDDI also uses a token to control the access to the network. However, it differs from token ring in many ways. FDDI uses light pulses to transmit information from one device to the next along the ring. In FDDI multiple devices can send several frames using timed token access method. This method allows any device with a token to transmit as many frames as needed in a fixed amount of time. It is called Token Rotation Time (TRT) and depends on the number of devices on the ring. More devices on the ring, less TRT allotted to each device and vice versa. So, FDDI can transmit more data at a faster speed. FDDI also uses dual counter rotating rings to transfer data, one is called primary ring and the other is secondary ring as shown in figure 8.3. Devices attached to dual ring FDDI networks are called Dual Attached Stations (DASs). Figure 8.3: A View of FDDI dual ring The secondary ring provides reliability to the FDDI network. It can be used either as an additional transmission path or backup path in case of primary ring failure. FDDI also uses beaconing to find fault and its location along the ring. Network ring connection is restored using auto-wrapping method. It joins both rings by isolating the faulty network device. Further information can be found at the following web sites: http://www-mm.informatik.uni-mannheim.de/veranstaltungen/animation/mac/fddi/http://www.iol.unh.edu/training/fddi/htmls/Wrap.html FDDI was initially designed to run on the fibre-optic cables but it can also run on copper (UTP/STP) cable using electrical signals called Copper Distributed Data Interface (CDDI). In FDDI/CDDI, all the Internetworking & Middleware 1 -Network Architectures (continued) 3/4 devices are connected to a central device called concentrator (similar to HUB or MAU). A CDDI-FDDI Translator allows interconnection of CDDI and FDDI networks. More information about how to use concentrators in CDDI/FDDI networks can be found at the following web site: http://www.cisco.com/univercd/cc/td/doc/product/cddi/c1100/con14icg/56153.htm The FDDI LAN standard is defined in ISO 9314. It can be used as a backbone technology to connect high-speed computers in a Local Area Network (LAN). FDDI is mainly used to interconnect LANs through a high speed and reliable backbone cable. Fibre network segments require two fibre cables, one for transmitting data, and one for receiving data. Each end of a fibre cable is fitted with a plug that can be inserted into a network adapter, hub, or switch. In the U.S., most cables use a square SC (Stick and Click) connector that slides and locks into place when inserted into a computers network card or connected to another fibre cable. European and some Asian countries use a round ST (Stick and Turn) connector instead. Figure 8.6: A View of FDDI Network Card Internetworking & Middleware 1 -Network Architectures (continued) 4/4