Router switch fabric protection using forward error correction

• US 20050083921A1
• Filed: 11/05/2004
• Published: 04/21/2005
• Est. Priority Date: 10/31/2000
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Chiaro Networks, Ltd.
• Inventors: Brewer, Tony M., Palmer, Gregory S., Shaw, Keith W., Kleiner, Jim, McDermott, Thomas C. III, Blackmon, Harry C., Walker, Dean E., Dozier, Harold W., Traylor, David
• Granted Patent: US 8,315,175 B2
• Time in Patent Office: Days
• US Class Current: 370/360
• CPC Class Codes: H04J 2203/0057 Operations, administration and maintenance [OAM]
• CPC Class Codes: H04L 2012/5627 Fault tolerance and recovery
• CPC Class Codes: H04L 49/1523 Parallel switch fabric planes
• CPC Class Codes: H04L 49/208 Port mirroring
• CPC Class Codes: H04L 49/254 Centralised controller, i.e. arbitration or scheduling
• CPC Class Codes: H04L 49/357 Fibre channel switches
• CPC Class Codes: H04Q 11/0005 Switch and router aspects
• CPC Class Codes: H04Q 11/0066 Provisions for optical burst or packet networks
• CPC Class Codes: H04Q 2011/0043 Fault tolerance

Abstract

Instead of alternatively utilizing only one fabric or the other fabric of a redundant pair, both fabrics simultaneously transmit duplicate information, such that each packet forwarding module (PFM) receives the output of both fabrics simultaneously. In real time, an internal optics module (IOM) analyzes each information chunk coming out of a working zero switch fabric; simultaneously examines a parallel output of a working one duplicate switch fabric; and compares on a chunk-by-chunk basis the validity of each and every chunk from both switch fabrics. The IOM does this by examining forward error correction (FEC) check symbols encapsulated into each chunk. FEC check symbols allow correcting a predetermined number of bit errors within a chunk. If the chunk cannot be corrected, then the IOM provides indication to all PFMs downstream that the chunk is defective. Under such conditions, the PFMs select a chunk from the non-defective switch fabric. Under error-free normal conditions, however, the PFMs select a chunk arbitrarily from a default switch fabric. In this way, each chunk in real time is selected from a non-defective source and is thus guaranteed to be error free. Accordingly, if a switch fabric fails, no information chunks are lost anywhere in the system.

Claims

1-67. -67. (canceled)

68. A method of switch fabric protection comprising:

simultaneously launching parallel duplicate data streams through two duplicated substantially identical switch fabrics, including launching a first data stream through a first switch fabric comprising a first N×

M crossbar switch, and launching a substantially identical second data stream through a second switch fabric comprising a substantially identical second N×

M crossbar switch, wherein N is the number of ingress ports and M is the number of egress ports of each respective N×

M crossbar switch;

receiving said parallel duplicate data streams after passing simultaneously through said first and said second switch fabric;

examining said received duplicate data streams in accordance with predetermined selection criteria;

if either of said duplicate data streams satisfies said criteria and the other said duplicate data stream does not satisfy said criteria, then selecting said duplicate data stream that satisfies said criteria and discarding said duplicate data stream that does not satisfy said criteria; and

if both of said duplicate data streams satisfy said criteria, then arbitrarily selecting one of said duplicate data streams and arbitrarily discarding the non-selected duplicate data stream;

wherein said switch fabrics transmit said data streams encapsulated in fixed sized chunk payloads, wherein said duplicate data streams comprise duplicate sequences of data structures, wherein said data structures are said fixed sized chunk payloads, wherein said data structures are encapsulated before said launching with error correction and detection code, and wherein said selecting of said data stream is performed on a structure-by-structure basis in accordance with said code encapsulated with said data structure.

69. The method of claim 68 wherein said error correction and detection code is stripped away from said data structure after said selecting.

70. The method of claim 68 wherein data delivery by said data streams is not interrupted by an occurrence selected from the group consisting of malfunction, failure, removal, and replacement of one of said two duplicated substantially identical switch fabrics.

71. The method of claim 68 wherein said examining is performed at an egress internal optics module interconnected with an egress port of each of said duplicated substantially identical switch fabrics

72. The method of claim 71 wherein said selecting and said discarding are performed at an egress packet forwarding module interconnected with said egress internal optics module.

73. The method of claim 72 wherein said encapsulated code is stripped away at an egress internal optics module.

74. The method of claim 68 wherein said data structures are encapsulated at an ingress internal optics module.

75. The method of claim 68 wherein N is equal to M.

76. The method of claim 68 wherein N is not equal to M.

77. The method of claim 68 wherein said crossbar switches are optical switches.

78. The method of claim 68 wherein said first and said second switch fabric are incorporated into a router system.

79. The method of claim 71 wherein said egress internal optics module is interconnected with said egress port through an optical fiber.

80. The method of claim 74 wherein said ingress internal optics module is interconnected with said ingress port through an optical fiber.

81. The method of claim 68 wherein said error correction and detection code corrects errors in said data structures.

82. The method of claim 68 wherein said error correction and detection code detects uncorrectable errors in said data structures.

83. A method of fault isolation and diagnostics in a switch fabric comprising:

launching a non-traffic-bearing data structure on a predetermined data path through said switch fabric from a first module interconnected with said switch fabric;

detecting and receiving said non-traffic-bearing data structure at a predetermined second module interconnected with said switch fabric;

examining said received non-traffic-bearing data structure in accordance with predetermined criteria;

if said non-traffic-bearing data structure satisfies said criteria, then determining that said predetermined data path is error-free; and

if said non-traffic-bearing data structure fails to satisfy said criteria, then determining that said predetermined data path is faulty.

84. The method of claim 83 wherein said switch fabric comprises multiple duplicated switch fabrics.

85. The method of claim 83 wherein said first module is selected from the group consisting of optical switch modules and internal optics modules.

86. The method of claim 83 wherein said second module is selected from the group consisting of optical switch modules and internal optics modules.

87. The method of claim 83 wherein said first module and said second module are the same module.

88. The method of claim 83 wherein said predetermined criteria comprise error correction and detection.

89. The method of claim 88 wherein said error correction and detection is forward error correction.

90. The method of claim 83 wherein said switch fabric comprises an optical crossbar switch.

91. The method of claim 83 wherein said predetermined data path comprises an optical fiber cable.

92. The method of claim 83 wherein said non-traffic-bearing data structure is a substantially fixed size diagnostic chunk.

93. The method of claim 92 wherein identifications representing physical identities of said respective first module and of said second module are encoded into said diagnostic chunk prior to said launching;

and

said predetermined criteria comprise coincidence between the physical identities of said respective first and second modules and said respective encoded identifications of said first and second modules.

94. The method of claim 83 wherein said switch fabric, said first module, and said second module are incorporated within a router system.

95. A communication network comprising:

two duplicated substantially identical switch fabrics, including a first switch fabric comprising a first crossbar switch and a second switch fabric comprising a substantially identical second crossbar switch, wherein each crossbar switch has a plurality of ingress ports and a plurality of egress ports;

wherein said first crossbar switch and said second crossbar switch are connected in substantially identical parallel data paths, such that each ingress port of said first crossbar switch and said second crossbar switch is interconnected with a data launching module, and each egress port of said first crossbar switch and said second crossbar switch is interconnected with a data receiving module;

wherein said data launching module is operable to encapsulate data in duplicate data streams with error correction and detection code; and

wherein said data receiving module is operable to examine said received duplicate data streams in accordance with the error correction and detection code;

if either of said duplicate data streams shows no uncorrectable errors and the other said duplicate data stream shows an uncorrectable error, then selecting said duplicate data stream that shows no uncorrectable errors and discarding said duplicate data stream that shows an uncorrectable error;

or

if both of said duplicate data streams show no uncorrectable errors, then arbitrarily selecting one of said duplicate data streams and arbitrarily discarding the non-selected duplicate data stream, wherein said selecting of said data stream is performed on a structure-by-structure basis in accordance with said error correction and detection code.

96. The system of claim 95 wherein the geometry of said switch fabrics is folded, such that a data launching module and a data receiving module occupy the same physical circuit card.

97. The system of claim 95 wherein the error correction and detection code is selected from the group consisting of forward error correction code and cyclic redundancy code.

98. The system of claim 95 wherein the data in the duplicate data streams is in fixed size chunk payloads.

99. The system of claim 95 wherein the two duplicated substantially identical switch fabrics are included in an optical router.