WO2001005102A1

WO2001005102A1 - Interconnecting network domains

Info

Publication number: WO2001005102A1
Application number: PCT/GB2000/001291
Authority: WO
Inventors: Martin Roderick Lea
Original assignee: Madge Networks Limited
Priority date: 1999-07-09
Filing date: 2000-04-06
Publication date: 2001-01-18
Also published as: GB9916173D0

Abstract

The present invention relates to a communications network having a number of network domains (1, 2, 3, 4) operating in accordance with respective communications protocols and a number of bridges (5, 6, 7) for interconnecting the network domains. Each bridge includes a number of interconnectable ports for coupling to the network domains, and a processor adapted to communicate with other bridges (5, 6, 7) on the network to allow an optimum path between domains to be determined. The optimum path is determined in accordance with path cost components which represent the ability of respective ports to transfer data.

Description

INTERCONNECTING NETWORK DOMAINS

The present invention relates to a bridge for interconnecting network domains of a communications network.

Point-to-point type Local Area Networks (LANs) typically consist of a number switching nodes or hub units which are interconnected so as to define a number of paths between the switching nodes or hub units. Communication between end stations coupled to such point-to-point type LANs is achieved by establishing a path through the network which links the two end stations. In the majority of these LANs, such paths are established and maintained by using either Transparent Bridging or Source Routing type protocols.

In transparent bridging, the point-to-point LAN is configured so that there can be only one path between any two points in the network. This is achieved using a "Spanning Tree Protocol" in which the switching nodes communicate to transfer status data indicating the current status of links in the network. Each switching node uses this information to determine a route through the network which links all the switching nodes. Any links not on the determined route are then disabled by having one of the switching nodes on the link block a respective input/output port .

The disabled links are used to provide redundancy should any of the active links fail. When such a fault occurs, the status data transferred between the switching nodes is updated accordingly. The switching nodes then determine a new route and the network is reconfigured.

Transfer of data between end stations is controlled by having each switching node maintain a record of which destinations are associated with each of the ports of the node. Accordingly, when a switching node first receives a data packet, it will determine the source address of the data and store this along with an indication of the port on which the data was received, in an address list.

The switching node then examines the destination address and compares this to the current address list. If the destination address matches a stored address, the data is transferred to the next switching node via the associated port. With only one route between any two points in the network, the stored address lists are used to define a path through the network for a given destination address. Accordingly, the data is transferred by transferring the data between nodes, until the intended destination is reached.

If the destination address does not match a stored address, a copy of the data is transferred from the switching node via every port, other than the one via which it was received. Accordingly, copies of the data will propagate through the network until a copy reaches its destination. When the destination end station generates response data which is transferred to the original source end station, this can follow the path that was defined by the data packet which successfully reached the destination.

In contrast to this, source routing is used on LANs which include multiple routes through the network between any two destinations. This is achieved by having the end stations store an indication of the intended path through the network in the data which is to be transferred. This is stored in a Route Information Field (RIF) in the data packet header.

The path is initially determined by having the transmitting end station generate an All Routes Explorer

(ARE) data packet which includes the address of a second destination end station but no routing information. This is transferred to a first switching node which modifies the packet by adding path data to the RIF including an indication of the address of the switching node, along with an indication of the ports via which the data packet was received and output. The data packet is then copied and transferred to each adjacent switching node via the remaining ports.

Each of these adjacent switching nodes similarly updates the path data in the RIF with its own address and input and output ports and transfers copies of the packet to further switching nodes. Eventually, the second end station will receive a number of copies of the packet, each copy having path data representing a respective path through the network, stored in the RIF. The second end station selects one of the data packets and hence the path defined therein. The second end station then generates a response data packet including a copy of the respective path data, which is transferred to the first end station along the defined path which is stored in the RIF. This is achieved by passing the data packet sequentially to each switching node indicated in the path data until it reaches the first end station.

Once the response data packet is received by the first end station, this can extract the path data and add it to any subsequent data packets that are transferred.

It is often desirable to connect together different types of LANs to form a larger communications network and this is generally achieved using a number of bridging devices. An example of such a communications network system is shown in Figure 1.

The communications network of Figure 1 includes first, second and third Local Area Networks (LANs) 1,2,3 which operate in accordance with a source routing protocol, such as Token Ring type LANs, and a fourth LAN 4 which operates in accordance with a transparent bridging type protocol, such as an Ethernet type LAN. As will be appreciated, each of the individual LANs 1,2,3,4 comprise a number of interconnected switching nodes which are configured to operate in accordance with the respective communications protocol.

The first, second and fourth LANs 1,2,4 are coupled to respective ports 5a, 5b, 5c of a bridge 5 and the third and fourth LANs are coupled to respective ports 7b, 7a of a bridge 7, as shown. As will be understood by a person skilled in the art, the ports 5a, 7a are adapted to communicate with a transparent bridging type LAN, whereas the ports 5b, 7b, 5c are adapted to communicate to with LANs that operate in accordance with the source routing type protocol .

Coupled to each LAN 1,2,3,4 is a respective communications network end station 8,9,10,11, although it will be realised that practically a number of network end stations would be coupled to each network.

The bridges 5,7 are designed to operate in accordance with a transparent bridging protocol and accordingly each bridge maintains an address list which is determined by monitoring via which port data is received for each end station 8,9,10,11. This information is stored in accordance with the MAC address of the end station enabling the bridge to transfer data out of the correct port on the basis of the MAC address of the end station 8,9,10,11. In use, each LAN will operate in accordance with its respective communications protocol, to allow communication between two end stations coupled to the LAN, as described above .

If data is to be transferred between two end stations coupled to different LANs which operate in accordance with the same communications protocol, for example the end stations 8,9, then communication is achieved as follows. If data has previously been transferred between the end stations, then the end station 8 generates data packets which are transferred across the first LAN 1, in accordance with route information stored in the RIF, to the port 5b. The bridge 5 then looks for the MAC address of the destination end station 9 in its address list which indicates which port 5a, 5b, 5c the data packet should be output from, which in this case is the port 5c.

If data has not previously been transferred, the end station 8 will generate an ARE data packet which will be copied to the bridge 5. The bridge 5 updates the address list to indicate that data to be transferred to the end station 8 should be transferred via the port 5b.

The bridge 5 then copies the data packet out of both ports 5a, 5c, although only the copy of the data packet transferred out via the port 5c will reach the intended destination. When the end station 9 responds, the response data packet is transferred via the second LAN 2 to the bridge 5. The bridge 5 examines the data packet and determines from this information that the destination end station is the end station 8, and accordingly, transfers the data packet via the port 5b to the first LAN 1.

If data is to be transferred between the end stations which are coupled to LANs operating in accordance with different communications protocols, for example end stations 8,11, then the end station 8 generates an ARE data packet in the normal way which is transferred via the first LAN 1 to the port 5b and hence to the bridge 5. The bridge 5 uses the address list to transfer the data packet to the port 5a. Again if the address list does not indicate the port via which the data should be transferred, then the ARE data packet is copied to the port 5b and the port 5a.

Because the fourth LAN 4, is not adapted to operate in accordance with the source routing protocol, it is necessary for the port 5a to remove the RIF from the ARE data packet . The RIF is stored in a memory along with an indication of the MAC address of the source end station 8. Any other modifications required to the data packet are also carried out before the ARE data packet is transferred onto the fourth LAN 4, where it is transferred in accordance with the transparent bridging protocol to the end station 11.

When the end station 11 responds, the end station generates a data packet which is transferred across the fourth LAN 4 to the port 5a of the bridge 5, in accordance with the transparent bridging protocol . The data packet will include the MAC address of the destination end station 8 and this is used to access the RIF which is stored in memory. The RIF is added to the packet header and any other modifications required to the data packet are also carried out before the data packet is transferred to the port 5b. The data packet is then transferred across the first LAN 1 in accordance with the path information stored in the RIF.

Whilst this system works for relatively simple network configurations, problems occur where there are multiple routes between the LANs, as shown for example in Figure 2. In this example an additional bridge 6 is provided with the second and third LANs being coupled to respective ports 6b, 6c, as shown.

In this case, there are two possible routes connecting the two end stations 9,10 (i.e. via the second LAN 2, the bridge 5, the fourth LAN 4, the bridge 7 and the third LAN 3, or alternatively, via the second LAN 2, the bridge 6 and the third LAN 3) . This causes the network to fail when, for example, the end station 9 first attempts to communicate with an end station 10.

This occurs because as the end station 9 generates an ARE data packet, a first copy is transferred via the bridge 6 and the third LAN 3, to the bridge 7, whilst a second copy is transferred via the bridge 5 and the fourth LAN 4, to the port 7a of the bridge 7. This second copy of the ARE data packet will not include a RIF as it has been transferred over the fourth LAN 4 in accordance with the transparent bridging protocol .

Accordingly, the bridge 7 receives an ARE data packet via the port 7a and the port 7b, both of which contain the MAC address of the end station 9 as the source end station. Accordingly, the bridge attempts to update the address list stored in memory to indicate that data to be transferred to the end station 8 should be transferred out of the port 7a and the port 7b. As this is not a possible scenario for a transparent bridging protocol, this causes a conflict within the bridge 7 causing the system to fail . It will also be appreciated that in these circumstances the bridges 5,6 will also receive two copies of the ARE data packet via the ports 5c, 5a and 6b, 6c respectively. According to the first aspect of the present invention we provide a communications network comprising: a number of network domains operating in accordance with respective communications protocols; a number of bridges for interconnecting the network domains, each bridge including a number of interconnectable ports for coupling to the network domains, and a processor adapted to communicate with other bridges on the network to allow an optimum path between domains to be determined, the optimum path being determined in accordance with path cost components which represent the ability of respective ports to transfer data.

According to a second aspect of the present invention we provide a bridge for use in a network according to the first aspect of the invention. According to a third aspect of the present invention we provide a method of configuring a communications network, the method comprising: dividing the network into a number of network domains, each network domain operating in accordance with a respective communications protocol; interconnecting the network domains using a number of bridges; and, causing each bridge to communicate with other bridges on the network to allow an optimum path between domains to be determined, the optimum path being determined in accordance with path cost components which represent the ability of respective ports to transfer data.

The present invention therefore provides a bridge, a communications network and a method of configuring a communications network which allows a network which is formed from several interconnected domains to operate. The method operates by causing bridges which connect the network domains to communicate with each other thereby allowing a single path between the network domains to be determined. By then disabling ports on respective ones of the bridges, this results in a single path being obtained throughout the communications network thereby overcoming the problem of conflicts caused by the transfer of ARE data packets around the network.

Typically the bridges communicate with each other using Inter-domain Bridge Protocol Data Units (IBPDUs) , each IBPDU including a bridge identifier representative of the bridge which generated the IBPDU. Alternatively however, broadcast data packets or any other suitable form of communication may be used.

Typically each IBPDU further includes a port identifier, port priority number and path cost component associated with a respective port of the bridge generating the IBPDU.

Preferably the network uses an Inter-domain Spanning Tree Protocol (ISTP) to determine the path. However any suitable protocol that allows a path to be established may be used.

Generally at least one of the network domain comprises a number of interconnected switching nodes, with a path through the domains being determined in accordance with a Spanning Tree Protocol (STP) , although any form of network domain may be used. It will be realised that it is therefore possible for the network to operate the STP and ISTP in parallel, without any interference between the two, although this is not essential . Typically data having a first format is transferred over a first network domain operating in accordance with a first communications protocol, and data having a second different format is transferred over a second network domain operating in accordance with a second communications protocol, the bridge or bridge (s) connecting the first and second domains being adapted to translate data between the first and second formats thereby allowing data to be transferred between the first and second network domains. However, the present invention may equally apply to the situation in which the network domains operate in accordance with the same protocol. In this case, the present invention would typically be used to configure a large network. If the standard STP were to be used for the entire network, this can result in problems with free configuration should a portion of the network fail . In particular, any such reconfiguration would take a substantial length of time to filter through the entire network. However, if the present invention is implemented by splitting the overall network up into a number of individual network domains then should a fault occur on any one network domain, the protocol run by that network domain would operate to reconfigure the domain independently from the entire network. This results in the reconfiguration of the individual network domain much more rapidly than the whole network can be reconfigured. Furthermore, should a fault occur with one of the bridges, then the ISTP will operate to reconfigure the interconnections between the network domains therefore again resulting in a rapid reconfiguration of the entire network.

Typically, the bridge includes a store which stores the bridge identifier and the path cost component associated with each respective port. In this case the port identifier and associated port priority of each port in the bridge are also usually stored in the store.

Preferably at least one port includes an interface for translating data received thereat between first and second formats, thereby allowing the bridge to couple network domains operating in accordance with respective first and second communications protocols. However, as mentioned above, this is not essential as the system may be operated on a network in which each network domain operates in accordance with the same network protocol . The bridge generally includes a transfer store which stores data received at one of the ports before transferring the data to one or more of the other ports. This allows temporary storage of any data to be transferred in order to prevent overloading of the switch. However, it will be realised that if data is transferred directly between the ports, the transfer store would not be required.

An example of the present invention will now be described with reference to the accompanying drawings, in which: -

Figure 1 shows a first communications network;

Figure 2 shows a modified version of the communications network shown in Figure 1; Figure 3 shows an example of an IBPDU according to the present invention; and,

Figure 4 shows an example of a bridge according to the present invention.

Operation of the network shown in Figure 2 in accordance with the present invention will now be described.

In order to prevent the problems of one of the bridges 5,6,7 determining that an end station is coupled to more than one of the ports 5a, 5b, 5c, 6a, 6b, 6c, 7a, 7b, 7c the present invention configures the LANs 1,2,3,4 and the bridges 5,6,7 such that only a single path links the networks 1,2,3,4 and in the present example, this is achieved by running an Inter-domain Spanning Tree Protocol (ISTP) . The ISTP is a modified version of the Spanning Tree

Protocol (STP) method described in the IEEE 802. ID standard for controlling bridging paths through a network. The only modification is that the ISTP uses IBPDUs instead of BPDUs, as explained in more detail below. To avoid problems caused by bridging loops in the network, the ISTP temporarily eliminates loops by disabling ports so that there is only one possible path for the transmission of data packets across the network. In general, this protocol also aims to create a path that is more efficient and typically has a higher bandwidth than alternative paths. In order to achieve this, each bridge 5,6,7 must be capable of generating Inter-domain bridge protocol data units (IBPDU) which are then used in the same way as standard bridge protocol data units (BPDU) of the standard STP. An example of a bridge suitable for operating in accordance with the ISTP is shown in Figure 4.

The example shown in Figure 4 is that of the bridge 5, although it will be appreciated that the bridges 6 and 7 are identical to the bridge shown in Figure 2.

The bridge has a bus 30 which is linked to each of the ports 5a, 5b, 5c. Also coupled to the bus 30 is a bridge processor 31 and a bridge memory 32 which stores the address list required by the bridge 5, along with topology data, details of bridge and port identifiers and bridge and port path cost components, which are required by the ISTP. The two ports 5b, 5c which are coupled to source routing networks 1,2,3 include a processor 36 and a buffer memory 37 which is used for temporarily storing the data before it is transmitted from the port 5b, 5c or when it is received by the port 5b, 5c and before it is processed. The port 5a, which is coupled to the fourth LAN 4 includes a processor 33, a transmit and receiver buffer memory 34 and a translation memory 35.

In use, data received in the port 5a is temporarily stored in the buffer memory 34, prior to transfer through the bridge 5. The processor 33 analyses the received data packet to determine the destination MAC address of the destination end station.

If a data packet has previously been received by the port 5a from the destination end station, then processor 33 uses the MAC address to access the RIF which is stored in the translation memory 35. The processor 33 then adds the RIF to the header of the data packet, as well as making any additional modifications required to the data packet before it is transferred to the bridge processor 31, via the bus 30.

The bridge processor 31 then transfers the data packet to the respective port 5b, 5c in accordance with the address list stored in the bridge memory 32, where it is again temporarily held in the respective buffer memory 37. The processor 36 then controls the transfer of the data onto the respective first or second LAN 1,2 in accordance with the respective RIF.

If a data packet has not been received from the intended destination end station, then the translation memory will not contain a corresponding RIF, and the bridge memory 32 will not contain a corresponding entry on the address list. Accordingly, the processor 33 makes any required modifications to the data packet before transferring it to the bridge processor 31. The bridge processor 31 then copies the data packet to both the ports 5b, 5c so that it can be transferred to the first and second LANs 1,2 where it is transferred to the intended destination end station as an ARE data packet, thereby allowing a route between the bridge 5 and the destination end station to be determined in accordance with the normal source routing protocol . If a data packet is received at one of the ports 5b, 5c it is temporarily stored in the buffer memory 37 before transfer via the bus 30 to the bridge processor 31. The bridge processor 31 determines from the address list which port the data packet should be output from based on the MAC address of the destination end station and transfers the data packet accordingly. If the address list does not include an entry for the given destination MAC address, then the data packet is copied to both ports 5a, 5b.

If the data packet is to be sent to a different source routing network, then the data packet is transferred via the bus 30 to the other port 5b, 5c and hence onto the respective first or second LAN 1,2. Alternatively, if the data is to be transferred to the fourth LAN 4, the data is transferred via the bus 30 to the processor 33 of the port 5a. The processor 33 temporarily stores the data in the buffer memory 34. The processor also removes the RIF from the data packet and stores this in the memory 25 on the basis of the MAC address of the source end station, in the standard manner described above. The data packet is then transferred to the fourth LAN 4. If one of the ports 5b, 5c receives a BPDU type data packet from one of the LANs 1,2,4 then the data packet is transferred via the bus 30 to the bridge processor 31. The bridge processor 31 will determine whether this is a standard BPDU or an IBPDU. If it is a standard BPDU, then the data packet is intended to be used only by the first, second, or third LAN 1,2,3 which generated the BPDU for internal configuration purposes, and accordingly the BPDU is discarded. However, if it is an IBPDU, then the bridge processor 31 operates to compare the information contained in the IBPDU to the data stored in the bridge memory 32 in accordance with the spanning tree protocol set out above.

It will be understood that the IBPDU and the BPDU contain different header information allowing the distinction to be made between the two types of data packet. Accordingly, when an IBPDU is transferred via the LAN 4 which is using the STP for internal configuration purposes, the switching nodes within the fourth LAN 4 will not detect the data packet as a BPDU and therefore will not respond to the data contained therein. Instead the network will transfer the data packet as a data packet to be transferred across the network.

As a result, this allows the bridges 5,6,7 to be configured using the ISTP without affecting the operation of any of the LANs 1,2,3,4.

Operation of the STP will now be described with reference to the network system shown in Figure 2.

Each bridge 5,6,7 is assigned a unique bridge identifier B5,B6,B7 which is based on the MAC address of the respective bridge, the bridge identifier incorporating an associated priority indicated by a priority number. The bridge having the highest priority, which is indicated by the lowest priority number, is designated as the root bridge, which in the present example is the bridge 5.

Each port 5a, 5b, 5c, 6a, 6b, 6c, 7a, 7b, 7c of each bridge 5,6,7 is assigned a unique port identifier, which incorporates a respective port priority. Each port also has an associated path cost component. The path cost components are representative of the port's ability to transfer data. Typically the path cost value is set by default to a pre-set value, but can be re-set by a user to a lower value so as to focus traffic on that particular port (or to a higher value to divert traffic away) . Thus a port having a higher bandwidth is assigned a lower cost indicating an easier transfer whereas a lower bandwidth port is assigned a higher path cost component. The path cost components are used to calculate an overall path cost indicating the total cost of transferring data to the root bridge 5.

For each bridge, the ports 6b, 7a which are closest to the root bridge 5 are used to forward data to the root bridge and these are therefore known as root ports.

The details of each bridge including bridge and port identifiers, bridge and port priorities and the like are stored in an internal memory of the respective bridge.

The path costs are used by the bridges 5,6,7 to determine a designated bridge for each LAN. The designated bridge is the bridge 5,6,7 having the lowest path cost for transferring data from the respective LAN to the root bridge 5. In the present example, the designated bridge is the bridge 7 for the LAN 10. The port 7a which couples the designated bridge to the respective LAN is known as the designated port . Any port 6b which is not a root port or a designated port is placed in a blocking mode. This prevents data being transferred via this port thereby removing any loops from the network topology. If identical path costs are determined from a LAN to the root bridge 5, via two different bridges, then the bridge having the highest priority is the designated bridge. If the bridges have the same priority, then the priority of the respective ports is used to determine the designated bridge and the designated port .

The root bridge 5 and designated bridge 7 are determined by having all the bridges 5,6,7 communicate with each other to determine details of respective path costs and priority information. This is achieved by transmitting IBPDUs between the bridges, and having each bridge maintain a record of the information contained therein. This is stored in the memory in the form of topology data which indicates the status of each port of the respective bridge along with an indication of the root bridge.

An example of such an IBPDU data packet is shown in Figure 3. This includes a bridge field 20, which indicates the bridge identifier B5,B6,B7 of the bridge sending the BPDU, a root field 21 which indicates the bridge identifier Bl of the root bridge 1, a port field 22, which indicates the port identifier of the port 5a, 5b, 5c, 6a, 6b, 6c, 7a, 7b, 7c with which the IBPDU is associated and a root path cost field 23 which indicates the path cost back to the root bridge 5 from the respective port. There are also additional fields indicated generally at 24, although these are not relevant for the purposes of the present description.

Initially, each bridge 5,6,7 assumes it is the root bridge, and accordingly, it generates an IBPDU inserting its own bridge identifier B5,B6,B7 in the root field 21. Similarly a respective port is identified in the port field 22, and a value of zero is inserted in the root path cost field 23, as the cost of transferring data from the bridge to itself is zero. The generated IBPDU is then transmitted to all the other bridges via the second and third LAN 2,3. Upon receipt of an IBPDU, each bridge will compare the priority of the bridge identifier B5,B6,B7 indicated in the root field to the priority of the bridge identifier B1,B2,B3 ,B4,B5,B6 of the root bridge indicated in the topology data. If the indicated root bridge has a higher priority than the bridge identified in the IBPDU, the bridge will discard the IBPDU. If no root bridge is indicated in the topology data, the bridge will compare the root bridge identifier indicated in the IBPDU with its own identifier and if its own identifier has the higher priority, the bridge will generate a new IBPDU placing its own bridge identity in the root field. This is then transmitted onto the network in preference to the received IBPDU. Thus for example, if the bridge 5 received an IBPDU from any other bridge, it would determine that the priority of its own bridge identifier B5 was greater than that of the other bridge identifiers B6,B7. Accordingly, any BPDU indicating any other bridge identifier B6,B7 in the root field would be discarded and replaced.

If however the bridge identifier in the root field 21 has a higher priority, then the bridge will update the topology data stored in the memory and generate a new IBPDU. The new IBPDU will include at least some of the topology details from the received IBPDU, along with the bridge's own bridge identifier in the bridge field 21. The newly generated BPDU is then transmitted to all the other bridges accordingly.

In order to determine the designated bridge B5,B6,B7 for a given LAN, the path cost indicated in the IBPDU of bridges coupled to the LAN are compared. The bridge having the lowest path cost is then selected.

Thus, in the present example, the bridge 6 will generate an IBPDU indicating the path cost of transferring data from the LAN 14 to the root bridge 5. This will be transmitted to the bridge 7 which will compare it to its own path cost and determine its own path cost as lower. Accordingly, the bridge 7 will generate a response IBPDU which is returned to the bridge 6. Upon receiving this response IBPDU, the bridge 6 will determine that it is not the designated bridge for the fourth LAN 4 and will accordingly block the port 6b. Both bridges 6,7 update the topology data accordingly.

This process is repeated throughout the network until all the bridges are configured such that there are no loops within the network. In addition to this, in order to be able to update the network topology to account for any failures in the network the topology data stored in the bridges must be updated. In order to do this, the root bridge is configured by the ISTP to generate an IBPDU at regular intervals (such as every two seconds) . The other bridges update their topology data in accordance with the information contained in these BPDUs (which often remain the same from one frame to the next) . If however the root bridge does not generate an IBPDU, or this is at least not received by a bridge, then the affected bridge or bridges wait for a predetermined time-out interval (typically 15 seconds) before generating their own IBPDUs thereby allowing an alternative network configuration to be determined.

An alternative to the embodiment set out above is for the bridges 5,6,7 to communicate with each other using a broadcast data packet instead of IBPDUs. In this case, instead of having the IBPDUs specifically addressed to the bridges 5,6,7 on the basis of the bridge identifier B5,B6,B7 each bridge would generate a broadcast data packet including the topology data which has been generated in the same way. The broadcast data packet would then be broadcast throughout the entire network. However, only the bridges 6,7 would be adapted to detect the broadcast data packet and this would therefore be ignored by the remainder of the network switching nodes.

Claims

1. A communications network comprising: a number of network domains operating in accordance with respective communications protocols; a number of bridges for interconnecting the network domains, each bridge including a number of interconnectable ports for coupling to the network domains, and a processor adapted to communicate with other bridges on the network to allow an optimum path between domains to be determined, the optimum path being determined in accordance with path cost components which represent the ability of respective ports to transfer data.

2. A communications network according to claim 1, wherein the bridges communicate with each other using Inter-domain

Bridge Protocol Data Units (IBPDUs) , each IBPDU including a bridge identifier representative of the bridge which generated the IBPDU.

3. A communications network according to claim 2, wherein each IBPDU further includes a port identifier, port priority number and path cost component associated with a respective port of the bridge generating the IBPDU.

4. A communications network according to any of the preceding claims, wherein the network uses an Inter-domain Spanning Tree Protocol (ISTP) to determine the path.

5. A communications network according to any of the preceding claims, wherein at least one of the network domains comprises a number of interconnected switching nodes, with a path through the domain being determined in accordance with a Spanning Tree Protocol (STP) .

6. A communications network according to any of the preceding claims, wherein data having a first format is transferred over a first network domain operating in accordance with a first communications protocol, and data having a second different format is transferred over a second network domain operating in accordance with a second communications protocol, the bridge or bridge (s) connecting the first and second domains being adapted to translate data between the first and second formats thereby allowing data to be transferred between the first and second network domains .

7. A communications network substantially as hereinbefore described with reference to the accompanying drawings .

8. A bridge for use in a network according to any of the preceding claims.

9. A bridge according to claim 8, for use in a network according to at least claim 2, the bridge including a store which stores the bridge identifier and the path cost component associated with each respective port.

10. A bridge according to claim 9, for use in a network according to at least claim 3, wherein the port identifier and associated port priority of each port in the bridge are stored in the store.

11. A bridge according to any of claims 8 to 10, at least one port including an interface for translating data received thereat between first and second formats, thereby allowing the bridge to couple network domains operating in accordance with respective first and second communications protocols .

12. A bridge according to any of claims 8 to 11, the bridge including a transfer store which stores data received at one of the ports before transferring the data to one or more of the other ports .

13. A bridge for interconnecting network domains of a communications network, the bridge being substantially as hereinbefore described with reference to the accompanying drawings.

14. A method of configuring a communications network, the method comprising: dividing the network into a number of network domains, each network domain operating in accordance with a respective communications protocol; interconnecting the network domains using a number of bridges; and, causing each bridge to communicate with other bridges on the network using Inter-domain Bridge Protocol Data Units (IBPDUs) to allow an optimum path between domains to be determined, the optimum path being determined in accordance with path cost components which represent the ability of respective ports to transfer data.

15. A method according to claim 14, wherein data having a first format is transferred over a first network domain operating in accordance with a first communications protocol, and data having a second format is transferred over a second network domain operating in accordance with a second communications protocol, the method further comprising causing the bridge or bridge (s) connecting the first and second domains to translate data between the first and second formats thereby allowing data to be transferred between the first and second network domains.

16. A method according to claim 14 or claim 15, the method comprising causing the bridges to communicate with each other using Inter-domain Bridge Protocol Data Units (IBPDUs) , each IBPDU including a bridge identifier representative of the bridge which generated the IBPDU.

17. A method according to claim 16, the method further comprising causing each IBPDU to further include a port identifier, port priority number and path cost component associated with a respective port of the bridge generating the IBPDU.

18. A method according to claim 14 or claim 15, the method comprising causing the network to use an Inter-domain Spanning Tree Protocol (ISTP) to determine the path.

19. A method of configuring a communications network substantially as hereinbefore described with reference to the accompanying drawings .