System and method for rapid fault isolation in a storage area network

• US 20020194524A1
• Filed: 06/13/2002
• Published: 12/19/2002
• Est. Priority Date: 06/15/2001
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Wiley, Stephen A., Cadieux, Christian, Munson, Wayne
• Granted Patent: US 7,058,844 B2
• Time in Patent Office: Days
• US Class Current: 714/4.5
• CPC Class Codes: H04L 1/0061 Error detection codes
• CPC Class Codes: H04L 1/0082 fields explicitly indicating existence of error in data being transmitted, e.g. so that downstream stations can avoid decoding erroneous packet; relays
• CPC Class Codes: H04L 1/22 using redundant apparatus to increase reliability
• CPC Class Codes: H04L 41/0659 by isolating or reconfiguring faulty entities

Abstract

A fault region identification system adapted for use in a network, such as a storage area network (SAN), includes logic and/or program modules configured to identify errors that occur in the transmission of command, data and response packets between at least one host, switches and target devices on the network. The system maintains a count at each of a plurality of packet-receiving components of the network, the count indicating a number of CRC or other errors that have been detected by each component. The error counts are stored with the time of detection. The system alters the EOF (end-of-file) delimiter for each packet for which an error was counted such that other components ignore that packet, i.e. do not increment their error counts for that packet. Link segments adjacent single- or multiple-device components of the network are identified as fault regions, based upon the error counts of those components.

Claims

1. A fault isolation system adapted for use in a computer network having a host in communication with a storage device configured to store information relating to a plurality of detected errors that occur in packets transmitted over the network, the system having a plurality of program modules configured to execute on at least one processor, the program modules including:

an error detection module configured to identify respective components within the network at which each of the plurality of detected errors occurs;

an error count module configured to increment an error count at each identified component where an error has occurred;

a packet ignore module configured to alter a given packet for which an error has been detected to indicate to components other than the identified component not to increment their error counts for the given packet; and

a link segment identification module configured to identify at least one link segment coupled to each identified component at which an error count is incremented.

2. The system of claim 1, wherein:

a first incremented error count corresponds to a first switch on the network coupled to the host; and

the identified link segment relating to the first incremented error count includes a link segment between the first switch and the host.

3. The system of claim 1, configured for use in a configuration of the network including a plurality of switches, wherein:

a first incremented error count corresponds to one of the plurality switches on the network; and

the identified link segment relating to the first incremented error count includes a link segment between the one switch and a switch at an adjacent position on the network.

4. The system of claim 1, wherein:

a first incremented error count corresponds to an addressed, single-device target on the network; and

the identified link segment relating to the first incremented error count includes a link segment between the addressed target and an adjacent component of the network.

5. The system of claim 1, wherein:

a first incremented error count corresponds to a write operation and corresponds to an addressed device of a multiple-device target on the network, where the addressed device is a first device in the target;

the identified link segment relating to the first incremented error count includes a link segment between the addressed device and a component of the network adjacent the multiple-device target.

6. The system of claim 1, wherein:

a first incremented error count corresponds to a write operation and corresponds to an addressed device of a multiple-device target on the network, where the addressed device is not a first device in the target;

the identified link segment relating to the first incremented error count includes at least one of a plurality of link segments between the addressed device and a component of the network adjacent the multiple-device target.

7. The system of claim 1, wherein:

a first incremented error count corresponds to a port of a switch adjacent an addressed, single-device target on the network; and

the identified link segment relating to the first incremented error count includes a link segment between the addressed target and the adjacent switch.

8. The system of claim 1, wherein:

a first incremented error count relates to a read operation from an addressed device of a multiple-device target on the network and corresponds to a port of a switch adjacent the multiple-device target, where the addressed device is a last device in the target; and

the identified link segment relating to the first incremented error count includes a link segment between the addressed device and the adjacent switch.

9. The system of claim 1, wherein:

a first incremented error count relates to a read operation from an addressed device of a multiple-device target on the network and corresponds to a port of a switch adjacent the multiple-device target, where the addressed device is not a last device in the target; and

the identified link segment relating to the first incremented error count includes at least one of a plurality of link segments between the addressed device and the adjacent switch.

10. The system of claim 1, wherein:

a first incremented error count corresponds to a component of the host; and

the identified link segment relating to the first incremented error count includes a link segment between the host and a component which is adjacent the host on the network.

11. The system of claim 10, wherein the adjacent component includes an addressed, single-device target for an operation relating to the first incremented error count.

12. The system of claim 10, wherein the adjacent component includes a switch on the network between the host and an addressed target for an operation relating to the first incremented error count.

13. The system of claim 1, wherein:

a first incremented error count corresponds to a component of the host and relates to a read operation of an addressed device which is a last device of a multiple-device target adjacent the host; and

the identified link segment relating to the first incremented error count includes a link segment between the host and the addressed device.

14. The system of claim 1, wherein:

a first incremented error count corresponds to a component of the host and relates to a read operation of an addressed device which is not a last device of a multiple-device target adjacent the host; and

the identified link segment relating to the first incremented error count includes at least one of a plurality of link segments between the addressed device and the host.

16. The method of claim 15, further including the steps of:

identifying a second component adjacent the first component on the network; and

identifying a link segment coupling the first and second components as the fault region.

17. The method of claim 15, further including the steps of:

identifying a multiple-device target as the first component; and

identifying at least one of a plurality of link segments between the multiple-device target and an adjacent second component as the fault region.

19. The computer program product of claim 18, wherein the first component of the network comprises a multiple-device component, and the fault region detection module is configured to identify the fault region as including a plurality of link segments of the multiple-device component.

15. A method for identifying a fault region in a network having a host processor-based system in communication with at least one target device and a plurality of switches, including the steps of:

identifying at least one error relating to transmission of a packet on the network;

generating an error count relating to the identified error and corresponding to a first component at which the error was identified;

generating an indicator relating to the packet configured to inhibit other components from generating error counts relating to the identified error.

18. A computer program product stored on a computer-usable medium, comprising a computer-readable program configured to cause a computer to control execution of an application to identify a fault region associated with at least one of a plurality of detected errors in a network, the computer-readable program including:

an error identification module configured to identify at least one error relating to transmission of an error packet on the network;

an error count module configured to maintain an error count relating to at least one error identified at each of a plurality of components on the network;

a packet delimiter module configured to modify a packet for which an error is detected at a first component of the network to inhibit other components from generating error counts relating to the identified error; and

a fault region detection module configured to identify at least one link segment adjacent the first component as the fault region.

20. A computer network, including:

a host including a processor and a host bus adapter;

error identification logic configured to identify at least one error relating to transmission of an error packet on the network;

error count logic configured to maintain an error count relating to at least one error identified at each of a plurality of components on the network;

packet delimiter logic configured to modify a packet for which an error is detected at a first component of the network to inhibit other components from generating error counts relating to the identified error; and

fault region detection logic configured to identify at least one link segment adjacent the first component as the fault region.

21. The computer network product of claim 20, wherein the first component of the network comprises a multiple-device component, and the fault region detection logic is configured to identify the fault region as including a plurality of link segments of the multiple-device component.