Recoverable cache preload in clustered computer system based upon monitored preload state of cache

• US 7,676,635 B2
• Filed: 11/29/2006
• Issued: 03/09/2010
• Est. Priority Date: 11/29/2006
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Sutter, Kevin William, Newport, William Thomas
• Primary Examiner(s): Elmore; Stephen C
• Application Number:
  US11/564,420
• Publication Number:
  US 20080126706A1
• Time in Patent Office: 1,196 Days
• Field of Search: 711/137, 711/118, 711/144, 711/156, 714/5
• US Class Current: 711/137
• CPC Class Codes: G06F 11/1658 Data re-synchronization of a redundant component, or initial sync of replacement, additional or spare unit
• CPC Class Codes: G06F 11/2023 Failover techniques
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 11/2046 where the redundant components share persistent storage G06F11/2043 takes precedence
• CPC Class Codes: G06F 12/0862 with prefetch
• CPC Class Codes: G06F 12/0866 for peripheral storage systems, e.g. disk cache

Abstract

An apparatus, program product and method monitor the state of a cache during a cache preload operation in a clustered computer system such that the monitored state can be used after a failover to potentially avoid the need to restart the cache preload operation from the beginning. In particular, by monitoring the state of the cache during a cache preload operation, any data that has been preloaded into a cache prior to a failure may be retained after a failover occurs, thus enabling the cache preload operation to continue from the point at which it was interrupted as a result of the failure.

Claims

1. A method of preloading a cache in a clustered computer system, the method comprising:

monitoring a preload state of a cache during preloading of the cache by a first member of a group;

in response to a failure in the first member, failing over to a second member of the group; and

selectively preloading the cache with the second member based upon the preload state of the cache after failing over to the second member.

2. The method of claim 1, wherein monitoring the preload state includes storing preload status information during preloading of the cache, and wherein selectively preloading the cache with the second member includes retrieving the preload status information after the failure in the first member.

3. The method of claim 1, wherein selectively preloading the cache with the second member includes skipping preloading of the cache if the preload state of the cache indicates that the cache was fully preloaded prior to failure in the first member.

4. The method of claim 1, wherein selectively preloading the cache with the second member includes restarting preloading of the cache based upon the preload state of the cache.

5. The method of claim 4, wherein restarting preloading of the cache is performed if the preload state of the cache indicates that the first member had not yet begun preloading the cache as of the failure in the first member.

6. The method of claim 4, wherein restarting preloading of the cache is performed if the preload state of the cache indicates that the cache cannot be recovered.

7. The method of claim 1, wherein selectively preloading the cache with the second member includes starting preloading of the cache by the second member from an intermediate point such that at least a portion of any preloading of the cache performed by the first member is retained in the cache.

8. The method of claim 7, wherein selectively preloading the cache with the second member further includes determining the intermediate point from the preload state.

9. The method of claim 1, further comprising updating the preload state after each of a plurality of preload operations performed by the first member.

10. A method of preloading a cache in a clustered computer system, the method comprising:

initiating preloading of a cache by a first member of a group;

in response to a failure in the first member prior to completing the preload of the cache and after at least a portion of the cache has been preloaded, failing over to a second member of the group; and

after failing over to the second member, retaining the preloaded portion of the cache and completing the preload of the cache with the second member.

11. The method of claim 10, further comprising:

monitoring a preload state of the cache during preloading of the cache by the first member;

retrieving the preload state of the cache after the failure in the first member; and

determining an intermediate point at which to start preloading the cache with the second member from the retrieved preload state.

12. The method of claim 11, further comprising updating the preload state after each of a plurality of preload operations performed by the first member.

13. A method of preloading a cache in a clustered computer system of a type including a group including first and second members, wherein the first member is configured to preload the cache, the method comprising:

accessing a preload state of the cache with the second member after a failure has occurred in the first member and a failover to the second member has been initiated; and

selectively preloading the cache with the second member based upon the preload state of the cache after failing over to the second member.

14. An apparatus, comprising:

at least one hardware-based processor; and

program code configured to be executed by the at least one processor to preload a cache in a clustered computer system of the type including a group including a plurality of members, the program code associated with one of the plurality of members and configured to, in connection with a failover initiated after a failure in another member of the group, access a preload state for the cache prior to initiating a preload of the cache to determine whether at least a portion of the cache has already been preloaded by the other member of the group.

15. The apparatus of claim 14, wherein the program code is further configured to selectively preload the cache based upon the preload state, and update the preload state during selective preloading of the cache.

16. The apparatus of claim 15, wherein the program code is configured to selectively preload the cache by skipping preloading of the cache if the preload state of the cache indicates that the cache was fully preloaded prior to failure in the other member.

17. The apparatus of claim 15, wherein the program code is configured to selectively preload the cache by restarting preloading of the cache based upon the preload state of the cache.

18. The apparatus of claim 17, wherein the program code is configured to restart preloading of the cache if the preload state of the cache indicates that the other member had not yet begun preloading the cache as of the failure in the other member.

19. The apparatus of claim 17, wherein the program code is configured to restart preloading of the cache if the preload state of the cache indicates that the cache cannot be recovered.

20. The apparatus of claim 15, wherein the program code is configured to selectively preload the cache by starting preloading of the cache from an intermediate point such that any preloading of the cache performed by the other member is retained in the cache.

21. The apparatus of claim 20, wherein the program code is configured to selectively preload the cache further by determining the intermediate point from the preload state.

22. The apparatus of claim 14, wherein the at least one processor and the program code are resident in a first node of the clustered computer system, the apparatus further comprising a second node upon which is resident the other member of the group, the second node including second program code configured to monitor the preload state of the cache during preloading of the cache by the other member of the group.

23. The apparatus of claim 22, wherein the second program code is configured to monitor the preload state by storing preload status information during preloading of the cache, and wherein the first program code is configured to access the preload state by retrieving the preload status information after failure in the other member.

24. The apparatus of claim 23, wherein the second program code is configured to update the preload state after each of a plurality of preload operations performed by the first member.

25. A program product, comprising:

program code configured upon execution to preload a cache in a clustered computer system of the type including a group including a plurality of members, the program code associated with one of the plurality of members and configured to, in connection with a failover initiated after a failure in another member of the group, access a preload state for the cache prior to initiating a preload of the cache to determine whether at least a portion of the cache has already been preloaded by the other member of the group; and

a recordable type computer readable medium storing the program code.