Abstract: A method and apparatus are presented for operating a time slicing shared memory switch. The apparatus includes a bus for receiving a plurality of data frames in a respective plurality of input channels to the switch. A slice crosspoint applies the plurality of data frames to a shared memory in a time sliced manner. The time slice is established for each section of a shared memory to be staggered so that on any clock cycle, one memory portion is being accessed for writing at least some of the data frames and on a next clock cycle the memory portion is accessed for reading at least a portion of the data.

Abstract: A data communication architecture including a plurality of devices having input/output (I/O) ports supporting communication at a first rate and a data processor having a number of I/O ports where each I/O port supports data communication at a second data rate. The second data rate is at least double the first data rate. A communication link coupled to one of the data processor I/O ports supports the second data rate. A bridge device is coupled to the communication link and to the I/O ports of the plurality of devices. The bridge device translates the communication link at the second data rate to a plurality of communication links at the first data rate, where the plurality of communication links at the first data rate are substantially independent of each other.

Abstract: A Fiber Channel circuit (150) has a Fiber Channel switch (152) that receives data frames on input ports (142) and writes the frames to a data storage device (158) from an output port (156) A port server (162) backs up the frames to a data storage backup device (160) by reading data volumes from the data storage device (158) and writing backup copies of at least some of the read data volumes to the data storage backup device (160) A data monitor (166) receives the frame data from the output port and applies it to the data storage device (158) The data monitor (166) also applies it to the port server (162) The port server (162) identifies frames in the volume that have been changed during the backup process. The identified frames are either reread from the data storage array (158), or from a memory (168) associated with the port server (162), and written to the data storage backup device (160).

Abstract: A non-disruptive, on-line testing and switchover method and apparatus in a high availability, fibre channel switching environment. In a network having an active switching element, a redundant switching element and a port, one aspect of the present invention provides for verifying a working data path from the port to the redundant switching element, and thereafter verifying a working control path to the redundant switching element. Both verification tests take place without interrupting operation on the network. In addition, another aspect of the present invention provides for recognizing a failure occurrence in the active switching element, and thereafter switching to the redundant switching element with the potential for minimal frame loss.

Abstract: A method for scoring queued frames 18 for selective transmission through a switch (12) includes providing one or more switches in a fibre channel fabric, particularly one or more fibre channel switches (12′). The method includes assigning an initial score (20) to the content (42) of the one or more frames (18) of data (26). The initial score (20) is adjusted by one or more alternative score components to determine one or more adjusted scores (22). The adjusted scores (22) are compared. The method also provides for selecting frames (18) having the highest adjusted scores (22), and transmitting through the switches (12) the frames (18) having the highest adjusted scores (22).

Abstract: A route caching design in a fiber channel switch for providing quick access to recently used D_ID and exit port combinations. The fiber channel switch has a plurality of ports, each are coupled to a central route look-up table. A cache is coupled to each port for storing D_ID to exit port association information received from the central route look-up table.

Abstract: A non-disruptive, on-line testing and switchover method and apparatus in a high availability, fiber channel switching environment. In a network having an active switching element, a redundant switching element and a port, one aspect of the present invention provides for verifying a working data path from the port to the redundant switching element, and thereafter verifying a working control path to the redundant switching element. Both verification tests take place without interrupting operation on the network. In addition, another aspect of the present invention provides for recognizing a failure occurrence in the active switching element, and thereafter switching to the redundant switching element with the potential for minimal frame loss.

Abstract: A method and apparatus for routing frames through a fibre channel fabric to make the most efficient possible use of redundant inter-switch links between neighboring switches. The inter-switch links may have different bandwidths. The flow between adjacent switches is monitored to determine various local usage statistics and periodically adjust routing tables to move data flows from congested links to lightly loaded links.

Abstract: A method and apparatus are presented for operating a time slicing shared memory switch. The apparatus includes a bus for receiving a plurality of data frames in a respective plurality of input channels to the switch. A slice crosspoint applies the plurality of data frames to a shared memory in a time sliced manner. The time slice is established for each section of a shared memory to be staggered so that on any clock cycle, one memory portion is being accessed for writing at least some of the data frames and on a next clock cycle the memory portion is accessed for reading at least a portion of the data.

Abstract: An in-band method/apparatus whereby a host is enabled to secure predetermined operational information relative to predetermined ports of a fiber channel switch. A set command is generated at the host and sent in-band to the switch. The information content of the set command defines the ports for which operational-parameters are to be monitored. The information content of the set command also defines which operational parameters are to be monitored. In response to receiving the set command, the switch establishes statistical counters for monitoring port operational parameters in accordance with received operational parameter identifiers. An accept signal is then sent in-band to the host, and a time period of port monitoring begins. After a predefined time period has expired, the host sends a read command in-band to the switch. The switch now generates a monitor record in accordance with the count content of the statistical counters that were established in response to the set command.

Abstract: A method for implementing a link level service in a computer network having a first port device and a second port device. Node identification data is stored in the second port device. A physical-layer communications coupling is provided between the first port device and the second port device which may be a point-to-point, loop, or switched circuit connection. The first port device sends a request node identification (RNID) message addressed to the second port device. The second port device creates an accept message and copies stored node identification data into the accept message. The second port device sends the accept message to the first port device.

Abstract: The present invention is a fiber channel switch employing a distributed queuing algorithm for interconnecting a plurality of devices (workstations, supercomputer, peripherals) through their associated node ports (N_ports) and employs a fabric having a shared memory coupled to a plurality of fabric ports (F_ports) through a bi-directional bus over which memory addresses, frame data and communications commands are transmitted. Each F_port includes a port controller employing a distributed queuing algorithm associated with a control network for communicating commands between the ports related to when and where frame transfers should be made, wherein the bi-directional bus provides an independent data network for access to the shared memory such that frames can be transferred to and from the shared memory in response to port controller commands.

Abstract: A switch including a plurality of input/output (I/O) ports and a switching element programmably coupling a first of the I/O ports with a second of the I/O ports. An analysis device is associated with the first I/O port measuring at least one data traffic parameter specific to data traffic between the first I/O port and the second I/O port.

Abstract: A method and apparatus which obviates the throughput limitations retries generated by busy conditions of conventional switches which are limited to but a single path between any two ports through the use of a “hunt group” concept whereby multiple paths can be provided between a participating group member and another port. These multiple paths may be associated in groups of, for example, eight, with each grouping of eight paths then comprising a “hunt group” whereby the first available member of the hunt group will be selected to enable the connection should the first be “busy”. The particular embodiment disclosed comprises a plurality of sequentially organized hunt groups starting on a boundary of eight associated with ports 0-7, 8-15, . . . through 248-255.

Abstract: A method and apparatus for the establishment of dynamic Enterprise System Connection ("ESCON") connections over a Fibre Channel connection allows a port state machine (or port module) to request dynamic ESCON connections from the exchange context of the Fibre Channel frame and to implement dynamic linking of the Fibre Channel exchanges to ESCON ports while also linking and monitoring the status of these connections for all subsequent frames associated on a particular exchange. In a particular embodiment, the method and apparatus provides the ability to establish dynamic connections through an ESCON Director switch based off of the Originator Exchange Identifier ("OX.sub.-- ID") of the FC-2 header and information contained in the FC-4 header of the Fibre Channel frame. Once a connection path has been established for a particular OX.sub.-- ID, that connection is maintained until the exchange is terminated.

Abstract: A switch having a plurality of input/output (I/O) ports and a crossbar device programmably coupling a first of the I/O ports with a second of the I/O ports. A plurality of port request controllers (PRCs) are coupled such that each PRC is associated with one of the I/O ports. A plurality of serial request busses are arranged such that each serial request bus couples each PRC with its associated port. A plurality of serial response busses are coupled such that each serial response bus coupling each PRC with its associated PRC. In operation, the serial request and response busses operate independently in a non-blocking fashion to process connection and clear requests in parallel.

Abstract: A method for implementing a link level service in a computer network having a first port device and a second port device coupled by a communication link. Prior to a link incident being reported, the first port device executes a link incident record registration (LIRR) ELS message addressed to the second port device. The second port device responds to the LIRR by adding an address of the first port device to a registration list of ports registered to receive link incident reports. The second port device also responds to the LIRR by sending an accept reply message addressed to the first port device. After a link incident is detected by the second port device, the second port device generates a link incident record comprising data describing the link incident. The second port device selects an address from the registration list sends a registered link incident record ELS message addressed to the selected address.

Abstract: A high bandwidth central memory controller utilizing a pipelined TDM bus such that each serial interface can sustain a bandwidth of up to 100 MByte/second for both the transmission and reception of variable length frames. Each port is assigned a fixed number of queues, a TDM slot number and the address routing for all other queues associated with the remaining ports at initialization, such that when a frame is received, the appropriate queue is determined from the addressing in the frame header and the initialized route tables. When the port's TDM slot for a memory request is active, a request for the output queue is made to the central memory controller if an "output queue available" indication is returned and the frame data is placed on the bus during the input port's data TDM slot. If the output queue is not available, the input port may either discard the received data frame or generate a busy/reject frame to be placed on one of its own output queues during its TDM data slot.

Abstract: A method and apparatus are presented for operating a time slicing shared memory switch. The apparatus includes a bus for receiving a plurality of data frames in a respective plurality of input channels to the switch. A slice crosspoint applies the plurality of data frames to a shared memory in a time sliced manner. The time slice is established for each section of a shared memory to be staggered so that on any clock cycle, one memory portion is being accessed for writing at least some of the data frames and on a next clock cycle the memory portion is accessed for reading at least a portion of the data.

Abstract: A method and apparatus for configuring a system that includes a plurality of interconnected components that each supports service parameters for communicating with other components in the system. A determination is made as to which components support service parameters that are compatible, and groups of components having compatible service parameters are identified. Adjacent components exchange information frames that identify their service parameters. Each component compares its service parameters with those of its adjacent components to determine whether they are compatible, updating its own service parameters if necessary. Any component that updates its service parameters issues another information frame. Thus, information frames are exchanged until it is determined which components support compatible service parameters, and what service parameters are to be used for communicating among those components. Additionally, a unique address is automatically assigned to every port in the system.