Abstract: In one embodiment, a method is described for detecting an operational failure between the network unit and an adjacent network unit in the stack; controlling the switching engine to redirect packets which would otherwise be sent from a particular port to the adjacent network unit to be forwarded from another port to be sent to a different network unit in the stack; and entering the switching engine into a bypass mode in response to control data indicating an operational failure between at least two other network units in the stack to cause packets to be forwarded without being re-directed by the switching engine.

Abstract: A network unit for use in a switching fabric includes multiple units collectively constituting a single network entity, each having ports for the reception and forwarding of data packets. The network unit has at least one fabric port for connection to a partner port on another one of the units by at least one link. The network unit is organized to send and receive via the at least one fabric port protocol packets which contain information on the path costs between said units in the fabric and to perform an algorithm to determine, on the basis of said information, routes for data packets within the fabric to other units of the fabric.

Abstract: In one embodiment, a method is described for detecting an operational failure between the network unit and an adjacent network unit in the stack; controlling the switching engine to redirect packets which would otherwise be sent from a particular port to the adjacent network unit to be forwarded from another port to be sent to a different network unit in the stack; and entering the switching engine into a bypass mode in response to control data indicating an operational failure between at least two other network units in the stack to cause packets to be forwarded without being re-directed by the switching engine.

Abstract: A network unit for use in a switching fabric includes multiple units collectively constituting a single network entity, each having ports for the reception and forwarding of data packets. The network unit has at least one fabric port for connection to a partner port on another one of the units by at least one link. The network unit is organized to send and receive via the at least one fabric port protocol packets which contain information on the path costs between said units in the fabric and to perform an algorithm to determine, on the basis of said information, routes for data packets within the fabric to other units of the fabric.

Abstract: A network unit for use in a switching fabric includes multiple units collectively constituting a single network entity, each having ports for the reception and forwarding of data packets. The network unit has at least one fabric port for connection to a partner port on another one of the units by at least one link. The network unit is organized to send and receive via the at least one fabric port protocol packets which contain information on the path costs between said units in the fabric and to perform an algorithm to determine, on the basis of said information, routes for data packets within the fabric to other units of the fabric.

Abstract: A distributed router composed of individual units connected by links in a routing fabric which appears to be a single network entity routes multicast packets by bridging to the final unit in the fabric. At the final unit the packet is replicated as necessary and routed. The scheme avoids multiple transmission of the same packet across the fabric.

Abstract: A multi-port network unit for use in a cascade system of network units sends from a cascade port a packet including a special header having a source port ID, a destination port ID and a ‘destination port known’ field. The port IDs identify both a unit and a port within a unit. A routing database is set up, optionally by a discovery protocol, in terms of ports and either destination units (within the cascade) or source units (within the cascade). The database includes a mesh table, indicating from which cascade port a packet with a known destination port ID should be forwarded, without needing a fresh look-up. The database also includes a multicast exclusion table which allows ingress of packets with source unit IDs matched to an ingress port. The scheme allows a general mesh type of cascade while dynamically preventing closed loops.

Abstract: A network unit for use in a switching fabric includes multiple units collectively constituting a single network entity, each having ports for the reception and forwarding of data packets. The network unit has at least one fabric port for connection to a partner port on another one of the units by at least one link. The network unit is organized to send and receive via the at least one fabric port protocol packets which contain information on the path costs between said units in the fabric and to perform an algorithm to determine, on the basis of said information, routes for data packets within the fabric to other units of the fabric.

Abstract: A network unit for use in a switching fabric stores for each network unit in the switching fabric a respective change identification number. In each unit this number is incremented on each change of state. The unit broadcasts protocol packets which indicates incremented change identification numbers; receives corresponding packets from the other units; and determines when the last change identification numbers in the table are the same for all the units. This allows a control based on whether all the units know that all the other units have been updated in response to all the changes of state in the fabric. The network unit can also participate in a routing algorithm, and particularly a shortest path algorithm, within a fabric that constitutes a single network entity.

Abstract: A distributed router composed of individual units connected by links in a routing fabric which appears to be a single network entity routes multicast packets by bridging to the final unit in the fabric. At the final unit the packet is replicated as necessary and routed. The scheme avoids multiple transmission of the same packet across the fabric.

Abstract: A distributed router composed of individual units connected by links in a routing fabric which appears to be a single network entity routes multicast packets by bridging to the final unit in the fabric. At the final unit the packet is replicated as necessary and routed. The scheme avoids multiple transmission of the same packet across the fabric.

Abstract: Network units such as switches for use in a cascaded stack are organised to provide a cascade connection in the form of a dual unidirectional connection so that, in its ordinary configuration, there is at least one and preferably more than one unidirectional ring for each direction around the cascade, each ring including a respective port on each unit. For each ring, each port on a unit is connected by a respective link to a corresponding port on the preceding unit and the following unit. The units provide a self-healing operation in the event of various kinds of operational failure. The self-healing operation includes loop-back of packets in units adjacent the failure and bypass of a packet switching process for other units. The units include control logic for passing control frames containing status information relating to the units and links between them and for co-operation with a CPU to control a switching engine to perform the self-healing operation in accordance with that status information.

Abstract: A cascaded stack of routing units each of which routes packets that it receives at its external ports and which conduct an election for a lead router that lends network and media access control addresses to the router interfaces of the other units in the stack. These other subordinate units have selected ancillary functions suppressed but otherwise have full routing capability. A packet entering the stack and requiring routing is routed by the unit that receives it and is thereafter bridged to its egress port.

Abstract: A network unit for use in a switching fabric stores for each network unit in the switching fabric a respective change identification number. In each unit this number is incremented on each change of state. The unit broadcasts protocol packets which indicates incremented change identification numbers; receives corresponding packets from the other units; and determines when the last change identification numbers in the table are the same for all the units. This allows a control based on whether all the units know that all the other units have been updated in response to all the changes of state in the fabric. The network unit can also participate in a routing algorithm, and particularly a shortest path algorithm, within a fabric that constitutes a single network entity.

Abstract: A distributed router composed of individual units connected by links in a routing fabric which appears to be a single network entity routes multicast packets by bridging to the final unit in the fabric. At the final unit the packet is replicated as necessary and routed. The scheme avoids multiple transmission of the same packet across the fabric.

Abstract: A link aggregator control which produces for a system in which the aggregator is located data units identifying the system and a partner identity for a partner system, and in response to a change from an old system identity to a new system identity produces data units identifying the new system identity and distinctively identifying the old system identity in place of the partner identity. Where the system identity and partner identity are each in the form of a globally administered unicast media access control the old system identity is represented in the form of a locally administered multicast address in place of the partner identity.

Abstract: A trunk connection to a stack of network switches is coupled to physical ports distributed among the switches. A switch has an internal logical port for a protocol engine. A packet received by way of the trunk at a port on a switch which does not host the protocol engine is sent over a cascade connection to the host switch with a tag that is interpreted within the host switch as if the packet cam from the logical port. The tagging and interpretation distinguishes between packets which might or might not be intended for the protocol engine and did not arrive by way of the trunk from packets which are intended for the protocol engine and have been received by way of the trunk.

Abstract: A multi-port network unit for use in a cascade system of network units sends from a cascade port a packet including a special header having a source port ID, a destination port ID and a ‘destination port known’ field. The port IDs identify both a unit and a port within a unit. A routing database is set up, optionally by a discovery protocol, in terms of ports and either destination units (within the cascade) or source units (within the cascade). The database includes a mesh table, indicating from which cascade port a packet with a known destination port ID should be forwarded, without needing a fresh look-up. The database also includes a multicast exclusion table which allows ingress of packets with source unit IDs matched to an ingress port. The scheme allows a general mesh type of cascade while dynamically preventing closed loops.

Abstract: A cascaded stack of routing units each of which routes packets that it receives at its external ports and which conduct an election for a lead router that lends network and media access control addresses to the router interfaces of the other units in the stack. These other subordinate units have selected ancillary functions suppressed but otherwise have full routing capability. A packet entering the stack and requiring routing is routed by the unit that receives it and is thereafter bridged to its egress port.

Abstract: Network units such as switches for use in a cascaded stack are organised to provide a cascade connection in the form of a dual unidirectional connection so that, in its ordinary configuration, there is at least one and preferably more than one unidirectional ring for each direction around the cascade, each ring including a respective port on each unit. For each ring, each port on a unit is connected by a respective link to a corresponding port on the preceding unit and the following unit. The units provide a self-healing operation in the event of various kinds of operational failure. The self-healing operation includes loop-back of packets in units adjacent the failure and bypass of a packet switching process for other units. The units include control logic for passing control frames containing status information relating to the units and links between them and for co-operation with a CPU to control a switching engine to perform the self-healing operation in accordance with that status information.