Managing recovery of a link via loss of link

• US 8,432,793 B2
• Filed: 03/19/2008
• Issued: 04/30/2013
• Est. Priority Date: 02/25/2008
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Errickson, Richard K., Helmer, Leonard W. Jr., Houston, John S.
• Primary Examiner(s): Shah, Chirag
• Assistant Examiner(s): Van, Jenkey
• Application Number:
  US12/051,630
• Publication Number:
  US 20090216923A1
• Time in Patent Office: 1,868 Days
• Field of Search: 370216-228, 370/242
• US Class Current: 370/225
• CPC Class Codes: G06F 2209/548 Queue
• CPC Class Codes: G06F 9/546 Message passing systems or structures, e.g. queues

Abstract

A computer program product, apparatus and method for managing recovery of a link in a multi-tasking multi-processor environment. An exemplary embodiment includes shutting off timers for a failed channel associated with the communications link, storing a loss of link condition in a data structure, disabling communications on the failed channel and sending an external notification of the loss of link condition.

Claims

1. A computer program product for managing recovery of a communications link in a multi-tasking multi-processor environment, the computer program product comprising:

a tangible non-transitory storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising;

shutting off timers for a failed channel associated with the communications link, the failed channel being associated with a physical link;

storing a loss of link condition in a data structure;

disabling communications on the failed channel;

establishing a queue pair out-of-band connection in a channel that is multiplexed across the physical link;

sending an external notification, via the queue pair out-of-band connection, of the loss of link condition;

resuming, via the queue pair out-of-band connection, data communications on the failed channel, by rebuilding and connecting the failed channel in firmware across the physical link, thereby leaving the physical link intact; and

handling channel path management via the queue pair, wherein the channel path management comprises a connection establishment with a remote system, heartbeat, and an invalidation of requests and other recovery actions.

2. The computer program product as claimed in claim 1 wherein the method further comprises determining if a request to the channel is local.

3. The computer program product as claimed in claim 2 wherein the method further comprises in response to the request being local, sending an external notification that a loss of link has occurred.

4. The computer program product as claimed in claim 1 wherein the external notification of the loss of link condition is sent to an object broker.

5. The computer program product as claimed in claim 1 wherein the external notification of the loss of link condition is sent to an I/O processor.

6. The computer program product as claimed in claim 1 wherein the method further comprises receiving a notification that resources for the channel have been cleaned up.

7. The computer program product as claimed in claim 1 wherein the method further comprises initializing control blocks for the channel associated with the communication link.

8. The computer program product as claimed in claim 7 wherein the method further comprises sending a notification to channel layers to initiate a hardware dependent initialization.

9. An apparatus for managing recovery of a communications link in a multi-tasking multi-processor environment, the apparatus comprising:

a processor performing a method stored in a non-transitory storage medium, comprising;

shutting off timers for a failed channel associated with the communications link, the failed channel being associated with a physical link;

storing a loss of link condition in a data structure;

disabling communications on the failed channel;

establishing a queue pair out-of-band connection in a channel that is multiplexed across the physical link;

sending an external notification, via the queue pair out-of-band connection, of the loss of link condition;

resuming, via the queue pair out-of-band connection, data communications on the failed channel, by rebuilding and connecting the failed channel in firmware across the physical link, thereby leaving the physical link intact; and

handling channel path management via the queue pair, wherein the channel path management comprises a connection establishment with a remote system, heartbeat, and an invalidation of requests and other recovery actions.

10. The apparatus as claimed in claim 9 wherein the method further comprises determining if a request to the channel is local.

11. The apparatus as claimed in claim 10 wherein the method further comprises in response to the request being local, sending an external notification that a loss of link has occurred.

12. The apparatus as claimed in claim 9 wherein the method further comprises:

initializing control blocks for the channel associated with the communication link; and

sending a notification to channel layers to initiate a hardware dependent initialization.

13. A method for managing recovery of a communications link in a multi-tasking multi-processor environment, the method comprising:

shutting off timers for a failed channel associated with the communications link, the failed channel being associated with a physical link;

storing a loss of link condition in a data structure;

disabling communications on the failed channel;

establishing a queue pair out-of-band connection in a channel that is multiplexed across the physical link;

sending an external notification, via the queue pair out-of-band connection, of the loss of link condition;

resuming, via the queue pair out-of-band connection, data communications on the failed channel, by rebuilding and connecting the failed channel in firmware across the physical link, thereby leaving the physical link intact; and

handling channel path management via the queue pair, wherein the channel path management comprises a connection establishment with a remote system, heartbeat, and an invalidation of requests and other recovery actions.

14. The method as claimed in claim 13 further comprising determining if a request to the channel is local.

15. The method as claimed in claim 14 further comprising in response to the request being local, sending an external notification that a loss of link has occurred.

16. The method as claimed in claim 13 wherein the external notification of the loss of link condition is sent to an object broker.

17. The method as claimed in claim 13 wherein the external notification of the loss of link condition is sent to an I/O processor.

18. The method as claimed in claim 13 further comprising receiving a notification that resources for the channel have been cleaned up.

19. The method as claimed in claim 13 further comprising initializing control blocks for the channel associated with the communication link.

20. The method as claimed in claim 19 further comprising sending a notification to channel layers to initiate a hardware dependent initialization.

21. The method as claimed in claim 13 further comprising emulating channel protocols in the firmware running over a standardized transport.

22. The method as claim in claim 13 further comprising mapping buffers to the queue pair out-of-band connection, the buffers comprising a first buffer for receiving a command response and a second buffer for receiving data requests.