Message forwarding backup manager in a distributed server system

• US 7,685,227 B2
• Filed: 11/10/2006
• Issued: 03/23/2010
• Est. Priority Date: 11/10/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Miller, Thomas J., Gerber, Robert H.
• Primary Examiner(s): Nguyen; Quang N.
• Assistant Examiner(s): Zong; Ruolei
• Application Number:
  US11/558,887
• Publication Number:
  US 20080114827A1
• Time in Patent Office: 1,229 Days
• Field of Search: 709/201, 715/744
• US Class Current: 709/201
• CPC Class Codes: G06F 11/2033 switching over of hardware resources
• CPC Class Codes: G06F 11/2041 with more than one idle spare processing component
• CPC Class Codes: G06F 11/2097 maintaining the standby controller/processing unit updated initialisation or re-synchronisation thereof G06F11/1658 and subgroups
• CPC Class Codes: H04L 41/00 Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
• CPC Class Codes: H04L 43/0817 by checking functioning

Abstract

A distributed server system may include a primary manager, one or more backup managers communicatively coupled to the primary manager, and one or more workers communicatively coupled to the one or more backup managers. The one or more backup managers may receive a command to backup the state of the distributed server system. The backup managers may forward work tasks to the one or more communicatively coupled workers. The backup managers may further aggregate responses received from the one or more communicatively coupled workers and send the aggregated response to the primary manager.

Claims

1. A distributed server computer system, comprising:

a primary manager operable to provide distributed server management software services including generating and managing an identifier associated with a state of the distributed server computer system and further including managing and distributing work tasks, wherein the primary manager associates the identifier and the state with an aggregate amount of work performed by the work tasks since the state was initialized, the primary manager being a computer system;

one or more backup managers communicatively coupled to the primary manager each operable to back up information as requested by the primary manager and forwarding at least one request in response to a request received from the primary manager; and

one or more workers communicatively coupled to the one or more backup managers each operable to receive at least one request forwarded from the primary manager by the one or more backup managers.

2. The distributed server computer system of claim 1, wherein each of the one or more backup managers is further operable to receive a response from the one or more workers.

3. The distributed server computer system of claim 1, wherein each of the one or more backup managers is further operable to aggregate each response received from the one or more workers.

4. The distributed server computer system of claim 1, wherein each of the one or more backup managers is further operable to aggregate each response received from the one or more workers, and further operable to substitute information regarding the connection status of a worker of the one or more workers in place of a response if the worker of the one or more workers is not responding.

5. The distributed server computer system of claim 1, wherein the computer system also comprises the one or more backup managers.

6. The distributed server computer system of claim 1, wherein the computer system also comprises the one or more workers.

7. The distributed server computer system of claim 1, wherein the primary manager is further operable to reassign any one of the one or more workers to a responsive backup manager of the one or more backup managers if the any one of the one or more workers is assigned to a non-responsive backup manager of the one or more backup managers.

8. The distributed server computer system of claim 1, wherein the primary manager is further operable to reassign work assigned to a non-responsive worker of the one or more workers to a responsive worker of the one or more workers.

9. A method for performance by a backup manager in a distributed server system, the method comprising:

receiving a request from a primary manager to begin closing a state of the distributed server system, the state corresponding to an identifier associated with the state;

sending a response to the primary manager indicating that the state corresponding to the identifier has been closed;

receiving a request from the primary manager to broadcast tasks associated with the identifier; and

broadcasting the tasks associated with the identifier to one or more workers, wherein the backup manager is operable to store the identifier and the state of the distributed server system corresponding to the identifier, and wherein the backup manager is further operable to replace and assume the function of the primary manager should the primary manager cease functioning, the backup manager being a computer system, the method being performed by the backup manager.

10. The method of claim 9, further comprising persisting the state of the distributed server system.

11. The method of claim 9, further comprising receiving a response indicating the tasks have been performed by the one or more workers.

12. The method of claim 9, wherein the broadcasting the tasks associated with the identifier to the one or more workers further includes sending a request to close the identifier associated with the state of the distributed server system.

13. The method of claim 9, further comprising receiving a response from the one or more workers including new work to be performed.

14. The method of claim 9, further comprising sending a response to the primary manager including one or more new work requests received from the one or more workers.

15. One or more device-readable storage media having device-executable instructions for performing, by a primary manager, steps comprising:

sending a command to one or more backup managers to close a state of a distributed server system, the state corresponding to an identifier associated with the state, wherein at least one of the one or more backup managers is operable to replace and assume the function of the primary manager should the primary manager cease functioning;

waiting a first predetermined time interval to receive a response from each of the one or more backup servers;

determining if the number of backup servers that responded is greater than a predetermined number; and

reassigning one or more workers assigned to one or more backup servers that did not respond to one or more backup servers that did respond, wherein tasks associated with the identifier are broadcast by the one or more backup severs to the one or more workers.

16. The one or more device-readable storage media of claim 15, further comprising waiting a second predetermined time interval for the one or more backup servers that did not respond to respond.

17. The one or more device-readable storage media of claim 15, further comprising receiving information from the one or more backup servers about the responsiveness of the one or more workers.

18. The one or more device-readable storage media of claim 15, wherein the backup manager forwards the command to close an identifier associated with the state of the distributed server system to the one or more workers.