Link extenders with error propagation and reporting

• US 6,870,814 B1
• Filed: 10/12/1999
• Issued: 03/22/2005
• Est. Priority Date: 10/12/1999
• Status: Expired due to Term

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Benedict, Melvin Kent, Watson, William J., Cutts, Richard W. Jr., Brown, David A., Garcia, David J., Bunton, William P., Krause, John C.
• Primary Examiner(s): Pezzlo, John
• Assistant Examiner(s): Mills, Donald L
• Application Number:
  US09/415,731
• Time in Patent Office: 1,988 Days
• Field of Search: 370/242, 370/243, 370/244, 370/245, 370/248, 370/252, 370/225, 370/217
• US Class Current: 370/245
• CPC Class Codes: H04L 12/64 Hybrid switching systems
• CPC Class Codes: H04L 41/0659 by isolating or reconfiguring faulty entities
• CPC Class Codes: H04L 41/12 Discovery or management of network topologies

Abstract

A link extender node is used to extend links between end nodes and/or routing nodes in a system area network. A connection includes a first link, coupling an end or routing node to a local port of a first link extender, a second link coupling the remote ports of first and second link extenders, and a third link coupling the local port of the second link extender to an end or routing node. The link extender includes link exception detection logic and transmits a this link bad command on the link generating the exception and transmits an other link bad command on the link not generating the exception.

Claims

1. A node device in a system area network (SAN) comprising:

a first port, coupled to a first node, with the first port including a receiver for receiving data from the first node and a transmitter for transmitting data to the first node;

a second port, coupled to a second node, with the second port including a receiver for receiving data from the second node and a transmitter for transmitting data to the second node; and

control logic coupled to both the first and second ports, and operable for

detecting an exception at the receiver of either of the first and second ports and for generating a this link bad (TLB) command to be transmitted by one of the first and second ports that detected the exception and an other link bad (OLB) command to be transmitted from the other of the first and second ports that did not detect the exception,

passing an OLB command received by the receiver at the first port to the second port for transmission by the transmitter at the second port;

converting a TLB command received by the receiver at the first port to an OLB command and passing the OLB command to the second port for transmission by the transmitter at the second port.

3. The node device of claim 1, wherein the node device is one of a repeater and a two-port router.

2. In a system area network including end-nodes and routing nodes connected by physical links and link extenders for coupling links from either end-nodes or routing nodes to serial cables, where the link extenders are coupled in a connection comprising a first node having a port, a first link extender having local and remote ports, a first link coupling the port of said first node and the local port of said first link extender, a second link extender having local and remote ports, a second link coupling the remote ports of said first and second link extenders, a second node having a port, and a third link coupling the local port of said second link extender to the port of said second node, with the links being duplex links and each port having a receiver and transmitter, a method for determining, at said first and second nodes the identity of a faulty link in the connection, said method comprising the steps of:

detecting a link exception indication, at the receiver of one of said ports, identifying a faulty link;

if said first link is the faulty link;

transmitting a TLB (this link bad) command from the transmitter of the local port of said first link extender;

transmitting an OLB (other link bad) command from the transmitter of the remote port of said first link extender;

receiving the OLB command at the receiver of the remote port of said second link extender;

transmitting an OLB command from the transmitter of the local port of the second link extender so that the first node receives a TLB command and the second node receives an OLB command to indicate that the first link is faulty;

if said second link is the faulty link;

transmitting TLB commands from the transmitters of the remote ports of said first and second link extenders;

transmitting OLB commands from the transmitters of the local ports of said first and second link extenders so that the first and second nodes both receive OLB commands indicating that the second link is faulty;

if said third link is the faulty link;

transmitting a TLB command from the transmitter of the local port of said second link extender;

transmitting an OLB command from the transmitter of the remote port of said second link extender;

receiving said OLB command at the receiver of the remote port of said first link extender; and

transmitting an OLB command from the transmitter of the local port of said first link extender so that the first node receives an OLB command and the second node receives a TLB command indicating that the third link is faulty.

4. An interface arrangement in a system area network, the arrangement comprising:

a first interface-node having

a first network-port coupled to a first network end-node, the first end-node being operable according to a first protocol, and

a first extension-port;

a second interface-node having

a second network-port coupled to a second network end-node, the second end-node being operable according to a second protocol different than the first protocol, and

a second extension-port; and

a link coupling the first extension-port of the first interface-node and the second interface-port of the second interface-node;

each of the interface nodes including control logic operable to

detect an exception at one of the network-port and the extension-port,

responsively generate a this link bad (TLB) command to be transmitted through the one of the network-port and the extension port that detected the exception, and

responsively generate an other link bad (OLB) command to be transmitted from the other of the network-port and the extension port that did not detect the exception.

5. The interface arrangement of claim 4, wherein the control logic of each of the first and second interface-nodes is further operable to:

detect an OLB command received by one of the network-port and the extension-port; and

pass the OLB command to the other of the network-port and the extension port that did not receive the OLB command for transmission thereof.

6. The interface arrangement of claim 4, wherein the control logic of each of the first and second interface-nodes is further operable to:

detect a TLB command received by one of the network-port and the extension-port;

convert the received TLB command into an OLB command; and

pass the OLB command to the other of the network-port and the extension port that did not receive the TLB command for transmission thereof.

7. The interface arrangement of claim 4, wherein at least one of the interface-nodes is one of a repeater and a two-port router.

8. A system area network comprising:

a first node;

a first link-extension device;

a first link coupling the first node and the first link-extension device;

a second link-extension device;

a second link coupling the first link-extension device and the second link-extension device;

a second node;

a third link coupling the second link-extension device and the second node; and

fault-isolation logic, at least indirectly communicating with one or more of the first and second nodes, operable to identify which of the first, second or third links exhibited a fault.

9. The system area network of claim 8, wherein the fault-isolation logic is operable to

determine, for each of the first and second node, which of a link exception, this_link_bad (TLB) command or an other_link_bad (OLB) command has been reported;

and

identify the first, second or third link as exhibiting a coupling-fault based upon commands reported by the first node and the second node, respectively.

10. The method of claim 9, wherein the fault-isolation logic is operable to use relationships represented via the following look-up table

First Node:

11. The system area network of claim 8, wherein:

the first and second nodes are network end-nodes; and

the first and second end-nodes are operable according to different protocols.