公开(公告)号：GB2268374A

公开(公告)日：1994-01-05

申请号：GB9213240

申请日：1992-06-23

Applicant: IBM

Inventor： JUDD IAN DAVID ,  BEER REGINALD

IPC: H04L12/56 ,  H04L12/58 ,  H04L12/28

Abstract: Described is a network addressing scheme in which a message sent from a source node to a destination node include a path address which defines the path over which the message should travel to reach the destination node. At each node between the source and destination, the path address is compared against a predetermined value, and on determining that the address and predetermined value are different, the node modifies the address before forwarding the message onto the next node. In a switch node having three or more ports, the identity of the output port is determined from the path address and a portion of the address is deleted before sending the message out on that output port. Also described is a method of configuring a network in which one or more initiator nodes are defined, the initiator nodes issuing query messages to an adjacent node which responds by sending the initiator details of the number of operational ports which are implemented in the adjacent node. The initiator node then issues query messages addressed to those nodes which are attached to the operational ports on the adjacent node. This continues until the initiator has walked through the entire network at which point the configuration process is complete.

公开(公告)号：GB2377140A

公开(公告)日：2002-12-31

申请号：GB0115969

申请日：2001-06-29

Applicant: IBM

Inventor： BEER REGINALD ,  CASHMAN PAUL NICHOLAS ,  HOOTON PAUL ,  JUDD IAN DAVID ,  MADDOCK ROBERT FRANK ,  NICHOLSON ROBERT BRUCE ,  WHYTE BARRY DOUGLAS

IPC: H04L12/437 ,  H04L45/28

Abstract: A method and apparatus for recovery from faults in a loop network (500) is provided. The loop network (500) has a plurality of ports (520, 530, 532, 534) serially connected with means for bypassing the ports (520, 530, 532, 534) from the loop network (500). A control device (522, 524) is provided with bypass control over at least one of the ports (530, 532, 534). A host means (502) sends a command to the control device (522, 524) at regular intervals and the control device (522, 524) has a counter which restarts a time period at the receipt of each command. If the time period expires, the control device (522, 524) activates the means for bypassing all the ports (530, 532, 534) under its control. The loop network (500) may have two loops (516, 518) with at least some of the ports (520, 530, 532, 534) common to both loops (516, 518).

公开(公告)号：GB2295032A

公开(公告)日：1996-05-15

申请号：GB9422932

申请日：1994-11-14

Applicant: IBM

Inventor： BEER REGINALD ,  CUTTS STANLEY JOHN ,  KEW STEVEN BRIAN ,  MORRIS NEIL ,  NEWMARCH DAVID

IPC: G05B23/02 ,  G11B33/14 ,  G11B33/10

Abstract: A cooling system provides redundant fault-tolerant cooling to electrical devices in a data storage system and comprises a plurality N+1 individual cooling means, N cooling means being required to provide cooling of said devices and further comprises monitor logic for monitoring the status of said plurality of cooling means and for issuing a non-critical warning to said apparatus in the event of failure of one of said cooling means, said logic being further operable, in the event of a failure in said logic, to issue a non-critical warning to said apparatus.

公开(公告)号：GB2252020A

公开(公告)日：1992-07-22

申请号：GB9026336

申请日：1990-12-04

Applicant: IBM

Inventor： SHAH VINAY VELJI ,  JUDD IAN D ,  BEER REGINALD

IPC: H04L1/18 ,  H04L1/16 ,  H04L29/06 ,  H04L29/08

Abstract: Described is a system and method of communicating between first and second nodes connected by a serial link wherein the data is transmitted between the nodes in the form of packets made up of multiple bit frames. Each packet that is correctly received by the second node is acknowledged by means of a pair of multibit frames. These frames may be interleaved among the frames making up any outgoing data packet that is being transmitted by the second node. The technique employed provides unambiguous acknowledgment of each data packet and if a data packet is received incorrectly, the packet is available in a buffer in the first node for resending.

公开(公告)号：GB2377140B

公开(公告)日：2005-01-19

申请号：GB0115969

申请日：2001-06-29

Applicant: IBM

Inventor： BEER REGINALD ,  CASHMAN PAUL NICHOLAS ,  HOOTON PAUL ,  JUDD IAN DAVID ,  MADDOCK ROBERT FRANK ,  NICHOLSON ROBERT BRUCE ,  WHYTE BARRY DOUGLAS

IPC: H04L12/437 ,  H04L45/28

Abstract: A method and apparatus for recovery from faults in a loop network (500) is provided. The loop network (500) has a plurality of ports (520, 530, 532, 534) serially connected with means for bypassing the ports (520, 530, 532, 534) from the loop network (500). A control device (522, 524) is provided with bypass control over at least one of the ports (530, 532, 534). A host means (502) sends a command to the control device (522, 524) at regular intervals and the control device (522, 524) has a counter which restarts a time period at the receipt of each command. If the time period expires, the control device (522, 524) activates the means for bypassing all the ports (530, 532, 534) under its control. The loop network (500) may have two loops (516, 518) with at least some of the ports (520, 530, 532, 534) common to both loops (516, 518).

公开(公告)号：GB2376612B

公开(公告)日：2004-06-09

申请号：GB0114614

申请日：2001-06-15

Applicant: IBM

Inventor： CASHMAN PAUL NICHOLAS ,  NICHOLSON ROBERT BRUCE ,  MADDOCK ROBERT FRANK ,  BEER REGINALD ,  MORRIS NEIL ,  JUDD IAN DAVID ,  WHYTE BARRY DOUGLAS ,  HOOTON PAUL

IPC: H04L12/24 ,  H04L12/437 ,  H04L12/42 ,  H04L12/26

Abstract: A method and apparatus are provided for fault location in a loop network ( 100, 200, 400 ). The network system having a host port ( 214 ) for supplying and receiving data and a plurality of successively connected ports ( 201, 202, 203, 204, 205 ) through which data from the host port ( 214 ) is transferred. A counter ( 122 ) for each port records data transfers in which the amount of data received at a destination port is less than an expected amount of data. When a transfer with less than the expected amount of data is identified for a data flow between a sending port ( 201 ) and a destination port ( 214 ), the counters are incremented for each port ( 202, 203, 204, 205, 214 ) after the sending port up to and including the destination port. Analysing means determines a fault location in the network system from the distribution of counts in the counters ( 122 ).

公开(公告)号：GB2384405A

公开(公告)日：2003-07-23

申请号：GB0201212

申请日：2002-01-19

Applicant: IBM

Inventor： BEER REGINALD ,  CASHMAN PAUL NICHOLAS ,  HOOTON PAUL ,  MORRIS NEIL ,  MADDOCK ROBERT FRANK ,  NICHOLSON ROBERT BRUCE

IPC: H04L12/42 ,  H04L12/423 ,  H04L12/437

Abstract: A method and apparatus for detection of a port name in a loop network is provided, particularly a loop network in the form of a Fibre Channel Arbitrated Loop (FC-AL). The loop network has a plurality of devices each device having at least one port on the loop network. The method includes determining which ports are populated with devices for which the unique port name (WWPN) is not known 302. The populated ports are then all bypassed 303 and a mode is entered on the loop network in which idle frames are transmitted around the loop network. One port is un-bypassed at a time and the port name from the un-bypassed port is received and recorded. The port name is received from the un-bypassed port in a Loop Initialisation Select Master (LISM) frame transmitted by the un-bypassed port.

公开(公告)号：DE69331868T2

公开(公告)日：2003-01-02

申请号：DE69331868

申请日：1993-05-18

Applicant: IBM

Inventor： JUDD IAN DAVID ,  BEER REGINALD

IPC: H04L12/56

Abstract: Described is a network addressing scheme in which a message sent from a source node to a destination node include a path address which defines the path over which the message should travel to reach the destination node. At each node between the source and destination, the path address is compared against a predetermined value, and on determining that the address and predetermined value are different, the node modifies the address before forwarding the message onto the next node. In a switch node having three or more ports, the identity of the output port is determined from the path address and a portion of the address is deleted before sending the message out on that output port. Also described is a method of configuring a network in which one or more initiator nodes are defined, the initiator nodes issuing query messages to an adjacent node which responds by sending the initiator details of the number of operational ports which are implemented in the adjacent node. The initiator node then issues query messages addressed to those nodes which are attached to the operational ports on the adjacent node. This continues until the initiator has walked through the entire network at which point the configuration process is complete.

公开(公告)号：GB2377144A

公开(公告)日：2002-12-31

申请号：GB0117092

申请日：2001-07-13

Applicant: IBM

Inventor： BEER REGINALD ,  CASHMAN PAUL NICHOLAS ,  JUDD IAN DAVID ,  HOOTON PAUL ,  MADDOCK ROBERT FRANK ,  MORRIS NEIL ,  NICHOLSON ROBERT BRUCE ,  WHYTE BARRY DOUGLAS

IPC: H04L12/437 ,  H04L45/28

Abstract: A method and apparatus for recovery from faults in a loop network (400) is provided. The loop network (400) has a host means (402), a first loop and a second loop (406, 408), a plurality of ports (410) connected to each of the loops (406, 408) and a control device (414, 440) on or connected to each loop (406, 408) with bypass control over at least one of the ports (410) connected to the loop (406, 408). In the event of a failure on the first loop (406), the host means (402) instructs the bypassing of at least one port (410) on the first loop (406), the host means (402) sending the instructions via the control device (414, 440) on or connected to the second loop (408). The host means (402) may determine the physical topology of the ports (410) on the first and second loops (406, 408) such that when a failure is reported to the host means (402) by a port (432) on the first loop (406), the host means (402) instructs the bypassing of a port (434) in a specific relationship to the reporting port (432) on the first loop (434). The host means (402) may instruct the bypassing of all ports (416) on the first loop (406) and subsequently selectively un-bypass ports (416) to ascertain the location of a faulty port or ports.

公开(公告)号：GB2268373A

公开(公告)日：1994-01-05

申请号：GB9213151

申请日：1992-06-20

Applicant: IBM

Inventor： JUDD IAN DAVID ,  BEER REGINALD

IPC: H04L1/00 ,  H04L1/18 ,  H04L12/28 ,  H04L69/40 ,  H04L1/16

Abstract: An error recovery method for use in an information communication system which comprises a plurality of nodes connected by links. Information is transferred between the nodes in frames of predefined types, including at least a first frame type used to transfer data and a second frame type used for error recovery. Each node has at least a first and a second mode of operation, in the first mode frames of both first and second types are accepted, in the second mode frames of the first type are discarded and only frames of the second type are accepted. A master node which controls error recovery is selected from amongst those nodes which can initiate transfers.