Use of unique XID range among multiple control processors

• US 7,076,689 B2
• Filed: 10/29/2002
• Issued: 07/11/2006
• Est. Priority Date: 10/29/2002
• Status: Expired due to Fees

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Brocade Communications Systems, Inc. (Broadcom, Inc.)
• Inventors: Atkinson, Michael W.
• Primary Examiner(s): LE, DIEU MINH T
• Application Number:
  US10/282,952
• Publication Number:
  US 20040083402A1
• Time in Patent Office: 1,351 Days
• Field of Search: 714/13, 714/11, 714/12, 714/23
• US Class Current: 714/11
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare
• CPC Class Codes: G06F 11/2038 with a single idle spare processing component

Abstract

A “high availability” system comprises multiple switches under the control of multiple control processors (“CPs”). One of the CPs is “active,” while the other CP is in a “standby” mode. Upon detection of a failure of the active CP, fail-over occurs to the standby CP. The standby CP re-issues all messages that the previously active CP had issued and which had not yet completed when the fail-over occurred. The newly active CP uses a different range of transaction identifiers (“XIDs”) than its predecessor CP. By using a different range of XID values, the newly active CP determines whether an incoming response message is responsive to a message that originated from the previously active CP or the newly activated CP. The currently active CP ignores all response messages that do not have an XID in its range, thereby ensuring proper and reliable messaging.

Claims

1. A method of failing over from a first processor to a second processor, each processor including a transaction identifier (“

XID”

) in requests issued by the processor, comprising;

(a) upon the occurrence of a fault associated with the first processor, the second processor determining whether the first processor has issued a request that has not yet completed;

(b) the second processor reissuing any such request that has not yet completed using a XID that is different than the XID that was used by the first processor;

(c) the second processor receiving a response to an issued request; and

(d) the second processor determining whether the response contains the XID pertaining to the first processor or the second processor.

2. The method of claim 1 further including ignoring data contained in the response if the response contains an XID pertaining to the first processor.

3. The method of claim 1 further including processing data contained in the response if the response contains an XID pertaining to the second processor.

4. The method of claim 1 further including ignoring data contained in the response if the response contains an XID pertaining to the first processor, but processing data contained in the response if the response contains an XID pertaining to the second processor.

5. The method of claim 1 further assigning each processor a unique, non-overlapping range of XIDs.

6. A system, comprising:

a plurality of electronic devices; and

a pair of control processors coupled to each other and to the electronic devices;

wherein one control processor is in an active mode (“

active control processor”

) and the other control processor is in a standby mode (“

standby control processor”

); and

upon a fault occurring in the active control processor, the standby control processor determines whether the active control processor issued a request that has not yet completed, reissues any such request that has not yet completed, said reissued requests including a transaction identifier (“

XID”

) that is different than the XID used by the active control processor, said standby control processor also receives a response to an issued request and determines whether said response contains an XID pertaining to the active control processor or the standby control processor.

7. The system of claim 6 wherein said standby processor ignores data contained in said response if said response contains an XID pertaining to the active control processor.

8. The system of claim 6 wherein said standby processor processes data contained in said response if the response contains an XID pertaining to the standby control processor.

9. The system of claim 6 wherein said standby processor ignores data contained in said response if the response contains an XID pertaining to the active control processor, but processes data contained in said response if the response contains an XID pertaining to the standby control processor.

10. The system of claim 6 wherein the active and standby control processors each is assigned a unique, non-overlapping range of XIDs.

11. The system of claim 6 wherein said electronic devices comprise network switches.

12. A standby control processor adapted to communicate with an active control processor and other electronic devices, said standby and active control processors capable of issuing requests containing a transaction identifier (“

XID”

), comprising;

a CPU;

memory coupled to said CPU; and

firmware stored in said memory and executed by said CPU, said firmware permitting said standby control processor to;

determine when a fault associated with the active control processor has occurred;

determine whether the active control processor issued a request that has not yet completed;

reissue any such request that has not yet completed using an XID that is different than the XID used by the active control processor;

receive a response to an issued request; and

determine whether said response contains an XID pertaining to the active control processor or to the standby control processor.

13. The standby control processor of claim 12 wherein firmware causes said standby processor to ignore data contained in said response if said response contains an XID pertaining to the active control processor.

14. The standby control processor of claim 12 wherein firmware causes said standby processor to process data contained in said response if the response contains an XID pertaining to the standby control processor.

15. The standby control processor of claim 12 wherein firmware causes said standby processor to ignore data contained in said response if the response contains an XID pertaining to the active control processor, but process data contained in said response if the response contains an XID pertaining to the standby control processor.

16. The standby control processor of claim 12 wherein said standby control processor includes a range of XIDs that is unique and does not overlap the XIDs used by the active control processor.

17. A computer readable storage medium for storing an executable set of software instructions that are executable by a CPU, said software instructions being operable to provide fail-over functionality from an active processor to a standby processor, said software instructions comprising:

a means for detecting a fault associated with the active processor;

a means for determining whether the active processor has issued a request that has not yet completed;

a means for reissuing any such request that has not yet completed using a XID that is different than the XID that was used by the active processor;

a means for receiving a response to an issued request; and

a means for determining whether the response contains the XID pertaining to the first processor or the second processor.

18. The computer readable storage medium of claim 17 wherein said software further includes a means for ignoring data contained in the response if the response contains an XID pertaining to the active processor.

19. The computer readable storage medium of claim 17 wherein said software further includes a means for processing data contained in the response if the response contains an XID pertaining to the standby processor.

20. The computer readable storage medium of claim 17 wherein said software further includes a means for ignoring data contained in the response if the response contains an XID pertaining to the active processor, but processing data contained in the response if the response contains an XID pertaining to the standby processor.

21. The computer readable storage medium of claim 17 wherein said software further includes a means for assigning each processor a unique, non-overlapping range of XIDs.

22. A method of failing over from a first processor to a second processor, each processor including a transaction identifier (“

XID”

) in requests issued by the processor, comprising;

(a) the second processor receiving a response to an issued request;

(b) the second processor determining whether the response contains the XID pertaining to the first processor or the second processor; and

(c) the second processor reissuing any issued request that has not yet completed using a XID that is different than the XID that was used by the first processor.

23. The method of claim 22 further including ignoring data contained in the response if the response contains an XID pertaining to the first processor.

24. The method of claim 22 further including processing data contained in the response if the response contains an XID pertaining to the second processor.

25. The method of claim 22 further including ignoring data contained in the response if the response contains an XID pertaining to the first processor, but processing data contained in the response if the response contains an XID pertaining to the second processor.

26. The method of claim 22 further assigning each processor a unique, non-overlapping range of XIDs.

27. A system, comprising:

a plurality of electronic devices; and

a pair of control processors coupled to each other and to the electronic devices;

wherein one control processor is in an active mode (“

active control processor”

) and the other control processor is in a standby mode (“

standby control processor”

), both control processors including transaction identifiers (“

XIDs”

) in requests they issue and the XIDs used by the active control processor being different than the XIDs used by the standby processor; and

wherein upon fail-over to said standby control processor, said standby control processor receives a response to an issued request and determines whether said response contains an XID pertaining to the active control processor or the standby control processor and wherein said standby control processor reissues any requests from said active control processor that have not completed using a different XID than the XID used by the active control processor.

28. The system of claim 27 wherein said standby processor ignores data contained in said response if said response contains an XID pertaining to the active control processor.

29. The system of claim 27 wherein said standby processor processes data contained in said response if the response contains an XID pertaining to the standby control processor.

30. The system of claim 27 wherein said standby processor ignores data contained in said response if the response contains an XID pertaining to the active control processor, but processes data contained in said response if the response contains an XID pertaining to the standby control processor.

31. A method of transitioning a control processor from a first state to a second state, the control processor including a transaction identifier (“

XID”

) in requests issued by the processor, the XIDs differing from the first state to the second state, comprising;

(a) receiving a response to an issued request;

(b) determining whether the response contains an XID pertaining to the first state or the second state;

(c) reissuing any issued request that has not yet completed using an XID that is different than the XID used in the first state.

32. The method of claim 31 further including ignoring data contained in the response if the response contains an XID pertaining to the first state.

33. The method of claim 31 further including processing data contained in the response if the response contains an XID pertaining to the second state.

34. The method of claim 31 further including ignoring data contained in the response if the response contains an XID pertaining to the first state, but processing data contained in the response if the response contains an XID pertaining to the second state.