Method and system for automatically regenerating data on-demand

• US 6,665,688 B1
• Filed: 12/23/1998
• Issued: 12/16/2003
• Est. Priority Date: 12/23/1998
• Status: Expired due to Term

• Current Assignee: CRAY Incorporated (Hewlett-Packard Enterprise Company)
• Original Assignee: CRAY Incorporated (Hewlett-Packard Enterprise Company)
• Inventors: Koblenz, Brian D., Callahan, Charles David II
• Primary Examiner(s): Le, Uyen
• Assistant Examiner(s): Chen, Te Yu
• Application Number:
  US09/221,029
• Time in Patent Office: 1,819 Days
• Field of Search: 707/8, 707/3, 707/10, 707/101-200, 717/3, 717/11, 717/5, 717/10, 717/116, 717/151, 717/145
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2322 using timestamps

Abstract

A replay method and system for monitoring the generating of a data set from input data sets and, when the data set is subsequently accessed, automatically regenerating the data set if the data set is out-of-date. The replay system only regenerates those input data sets that are determined to be out-of-date and only regenerates the output data set if it is determined to be out-of-date. A data set is determined to be out-of-date only when an input data set has actually changed since the data set was last generated.

Claims

1. A method in a computer system for updating data sets, the method comprising:

when generating a data set, monitoring the generation of the data set and recording an identification of a service used to generate the data set and an identification of input data sets for the generation of the data set; and

when the data set is accessed,

determining whether the input data sets are up-to-date;

when an input data set is not up-to-date, regenerating the input data set that is not up-to-date using a service whose identification was recorded when the input data set was generated and indicating if the input data set has changed; and

when an input data set has been changed since the data set was last generated, regenerating the data set by using the identified service and the identified input data sets to generate an up-to-date data set.

2. The method of claim 1 wherein the input data sets have recorded identifications of services to be used to regenerate the input data sets when they are not up-to-date.

3. The method of claim 2 wherein the identifications of services are recorded in replay objects.

4. The method of claim 3 wherein the data set includes a pointer to the replay object.

5. The method of claim 1 wherein each input data set has a time stamp indicating when it was last changed and wherein the data set is regenerated only when its time stamp is less than a time stamp of an input data set.

6. The method of claim 1 wherein an input data set represents a sub-division of an aggregate data set so that the input data set is changed only when that sub-division of the aggregate data set is changed.

7. The method of claim 1 wherein an aggregate data set is sub-divided into subset data sets so that each subset data set is updated less frequently than the aggregate data set.

8. The method of claim 1 wherein the service is for translating a computer program from one format to another.

9. The method of claim 1 wherein the identifiers are recorded in a replay object having a time stamp and each input data set has a time stamp and wherein the data set is regenerated when the input data sets are up-to-date and the time stamp of an input data set is later than a time stamp of the replay object.

10. The method of claim 9 wherein the time stamp of an input data set is updated only when the input data set changes.

11. The method of claim 9 wherein the time stamp of the replay object is updated whenever the data set is regenerated.

12. The method of claim 9 wherein the time stamp is not updated when the data set is regenerated, but the regenerated data set is the same as before being regenerated.

49. The method of claim 1 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set and the input data sets.

50. The method of claim 1 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set, the service, and the input data sets.

13. A method in a computer system for updating data sets, the method comprising:

for each of a plurality of data sets, providing an association of the data set, a service, and an input data set by monitoring that the service generates the data set using the input data set as input; and

when a data set is to be accessed,

determining whether the input data set for the data set is up-to-date;

when it is determined that the input data set for the data set is not up-to-date, regenerating the input data set for the data set using the service and the input data set for the input data set being regenerated to generate an up-to-date input data set for the data set; and

when the input data set for the data set has been regenerated since the data set was last generated, regenerating the data set using the service and input data set for the data set to generate an up-to-date data set.

14. The method of claim 13 wherein an executing service provides the association of a data set that it generates and an input data set used by the service.

15. The method of claim 13 wherein data sets form a hierarchy of data sets and wherein data sets lower in the hierarchy are used by services to regenerate data sets higher in the hierarchy.

16. The method of claim 13 wherein data sets form a hierarchy of data sets and wherein when one data set in the hierarchy is accessed, determining whether data sets lower in the hierarchy are up-to-date and regenerating data sets that are not up-to-date, and regenerating the one data set only when its input data set has changed as a result of the regeneration of the data sets that are determined not to be up-to-date.

17. The method of claim 13 wherein a data set is regenerated only when its input data set has changed since the data set was last regenerated.

18. The method of claim 13 wherein the association of a data set and a service is provided to a replay object along with a time stamp.

19. The method of claim 18 wherein the time stamp is updated when the data set changes as a result of being regenerated.

20. The method of claim 18 wherein the time stamp is not updated when the data set does not change as a result of being regenerated.

21. The method of claim 13 wherein the providing includes providing an argument to be used by the service to regenerate the data set.

51. The method of claim 13 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set and the input data sets.

52. The method of claim 13 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set, the service, and the input data sets.

22. A computer-readable medium including instructions for controlling a computer system to ensure that a data set is up-to-date, by a method comprising:

when each of a plurality of data sets is generated, associating with the data set one or more direct input data sets that are directly used to generate the data set by monitoring the generation of the data sets; and

when a data set is to be accessed,

regenerating each indirect input data set that is indirectly used to generate the data set to be accessed when it is determined that a direct input data set associated with that indirect input data set has changed; and

regenerating the data set to be accessed when a direct input data set associated with the data set to be accessed has changed as a result of regenerating an indirect input data set of the data set to be accessed.

23. The computer-readable medium of claim 22 wherein an executing service provides the associating of a data set that it generates with one or more direct input data sets used by the service to directly generate the data set.

24. The computer-readable medium of claim 22 wherein a data set is regenerated only when a direct input data set has changed since the data set was last regenerated.

25. The computer-readable medium of claim 22 wherein the association of a data set to one or more direct input data sets is provided to a replay object along with a time stamp.

26. The computer-readable medium of claim 25 wherein the time stamp is updated when the data set changes as a result of being regenerated.

27. The computer-readable medium of claim 25 wherein the time stamp is not updated when the data set does not change as a result of being regenerated.

28. The computer-readable medium of claim 22 wherein the associating includes providing an argument to be used by a service to regenerate the data set.

53. The computer-readable medium of claim 22 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set and the input data sets.

54. The computer-readable medium of claim 22 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set, the service, and the input data sets.

29. A computer-based replay system comprising:

means for monitoring the generation of a data set and storing an association between the data set and a service, and an input data set used to generate the data set;

means for determining whether the data set is up-to-date;

means for regenerating the input data set is up-to-date to bring it up-to-date; and

means for executing the service to regenerate the data set using the input data set when the input data set has been regenerated as a result of ensuring that the input data set is up-to-date.

30. The replay system of claim 29 wherein the means for executing records a time stamp indicating when the data set has changed and wherein it is determined the input data set has changed when its time stamp is later than the time stamp of the data set.

31. The replay system of claim 29 wherein the means for determining receives an argument that is used to regenerate the data set.

32. The replay system of claim 29 wherein the means for determining is based on an indication of the service when it generates the data set.

33. The replay system of claim 29 including means for identifying a replay object for the input data set.

34. The replay system of claim 33 wherein the means for identifying the replay object includes a pointer to the replay object.

55. The replay system of claim 29 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set and the input data sets.

56. The replay system of claim 29 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set, the service, and the input data sets.

35. A method in a computer system for generating data sets, the method comprising:

providing by monitoring the generation of data sets an indication that a first data set is generated by a first service with a second data set as input and an indication that the second data set is generated by a second service with a third data set as input;

determining whether the second data set is up-to-date;

when it is determined that the second data set is not up-to-date, regenerating the second data set using the second service with the third data set as input; and

when the second data set has been regenerated since the first data set was last generated, regenerating the first data set using the first service with the second data set as input.

36. The method of claim 35 wherein the determining of whether a data set is up-to-date includes determining whether input data sets that are directly or indirectly used to generate the data are up-to-date.

37. The method of claim 36 wherein the second data set represents a subdivision of an aggregate data set so that the second data set is regenerated only when that subdivision of the aggregate data set is changed.

38. The method of claim 37 wherein the aggregate data set is sub-divided into subset data sets so that each subset data set is updated less frequently than the aggregate data set.

39. The method of claim 35 wherein the first service is a service for translating a computer program from one format to another.

40. The method of claim 35 wherein the regenerating of the first data set is performed under control of a first replay object.

41. The method of claim 40 wherein the regenerating of the second data set is performed under control of a second replay object.

57. The method of claim 35 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set and the input data sets.

58. The method of claim 35 wherein the monitoring includes receiving from the service that generates the data set an indication of the data set, the service, and the input data sets.

42. A computer system for generating data sets, comprising:

a component that monitors the generation of a first data set and stores an indication that the first data set is generated by a first service with a second data set as input and that monitors the generation of the second data set and stores an indication that the second data set is generated by a second service with a third data set as input;

a component that determines whether the second data set is up-to-date and, when it is determined that the second data set is not up-to-date, regenerates the second data set using the second service with the third data set as input; and

a component that regenerates the first data set using the first service with the second data set as input when the second data set has changed since the first data set was last generated.

43. The computer system of claim 42 wherein the determining of whether a data set is up-to-date includes determining whether input data sets that are directly or indirectly used to generate the data are up-to-date.

44. The computer system of claim 43 wherein the second data set represents a sub-division of an aggregate data set so that the second data set is regenerated only when that subdivision of the aggregate data set is changed.

45. The computer system of claim 44 wherein the aggregate data set is subdivided into subset data sets so that each subset data set is updated less frequently than the aggregate data set.

46. The computer system of claim 42 wherein the first service is a service for translating a computer program from one format to another format.

47. The computer system of claim 42 wherein the regenerating of the first data set is performed under control of a first replay object.

48. The computer system of claim 47 wherein the regenerating of the second data set is performed under control of the second replay object.