System and method for providing increased database fault tolerance

• US 7,487,188 B2
• Filed: 07/25/2005
• Issued: 02/03/2009
• Est. Priority Date: 09/07/2004
• Status: Active Grant

• Current Assignee: Computer Associates Think Inc. (Broadcom, Inc.)
• Original Assignee: Computer Associates Think Inc. (Broadcom, Inc.)
• Inventors: Eaton, Kenneth W., Shuma, Kevin P., Lynn, Joseph B.
• Primary Examiner(s): Wong; Leslie
• Application Number:
  US11/189,376
• Publication Number:
  US 20060053183A1
• Time in Patent Office: 1,289 Days
• Field of Search: 707/202, 707/204, 714/6
• US Class Current: 707/687
• CPC Class Codes: G06F 11/1482 by means of middleware or OS functionality
• CPC Class Codes: G06F 11/2038 with a single idle spare processing component
• CPC Class Codes: G06F 2201/80 Database-specific techniques
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

In one embodiment, a Database Management System (DBMS) providing increased fault tolerance includes a first DBMS instance communicably coupled with a data store. The DBMS further includes a second DBMS instance communicably coupled with the data store, with the second DBMS instance including a shadow instance operable to monitor data processing of the first DBMS instance using a communication facility.

Claims

1. A system for providing increased fault tolerance in a data storage environment, comprising:

a first Database Management System (DBMS) instance that initiates at least one first process associated with execution of a request to interact with a data store, the first DBMS instance being a primary handler of requests to interact with the data store; and

a second DBMS instance that initiates at least one second process associated with the at least one first process, the second DBMS instance being a shadow instance that monitors the first DBMS instance, wherein the first DBMS instance and the second DBMS instance are communicably coupled with the data store, wherein the first DBMS instance is communicably coupled to the second DBMS instance through a communication facility, and wherein the first DBMS instance uses the communication facility to notify the second DBMS instance of the at least one first process,

wherein the at least one first process comprises execution of a first command to open a dataset in the data store,

wherein notification by the first DBMS instance to the second DBMS instance comprises communicating a message to the second DBMS instance, the message notifying the second DBMS instance of the dataset to be opened,

wherein the at least one second process comprises execution of a second command to open the dataset based, at least in part, on the message,

wherein the dataset to be opened is opened in parallel by the first DBMS instance and the second DBMS instance, whereby the dataset to be opened is opened by the second DBMS instance before failure of the first DBMS instance.

2. The system of claim 1, wherein the first DBMS instance further:

executes a data processing query using the opened dataset; and

communicates information associated with the data processing query to the second DBMS instance through the communication facility.

3. The system of claim 2, wherein:

the first DBMS instance generates a log comprising information related to data processing of the first DBMS instance; and

the second DBMS instance monitors data processing of the first DBMS instance using the communication facility to scan the log.

4. The system of claim 3, wherein the second DBMS instance automatically performs data processing in response to a failure of the first DBMS instance.

5. The system of claim 4, wherein the second DBMS instance:

reprocesses a pipeline executed by the first DBMS instance on the data store based, at least in part, on the scanned log; and

in response to identifying a rollback, schedules the rollback for the data store.

6. The system claim 2, the first DBMS instance executing a command to close the dataset and communicating a message to the second DBMS instance, the message notifying the second DBMS instance of the dataset to be closed;

and

the second DBMS executing a second command to close the dataset based, at least in part, on the message.

7. The system claim 1, wherein:

the second DBMS instance receives notification that the first DBMS instance failed to execute the request; and

the second DBMS instance attempts to execute the request using, at least in part, the at least one second process.

8. The system of claim 1, wherein the second DBMS instance comprises a local DBMS instance.

9. The system of claim 8, further comprising a third DBMS instance communicably coupled to the data store, the third DBMS instance comprising a second shadow instance operable to monitor data processing of the first DBMS instance using the communication facility and the third DBMS instance comprising a remote DBMS instance.

10. The system of claim 1, wherein the second DBMS instance comprises a remote DBMS instance.

11. The system of claim 1, wherein the first DBMS instance executes in a first logical partition and the remote second DBMS instance executes in a second logical partition.

12. A method for providing increased database fault tolerance in a data storage environment, wherein a first Database Management Service (DBMS) instance and a second DBMS instance are communicably coupled with a data store, and wherein the first DBMS instance is communicably coupled to the second DBMS instance through a communication facility, the first DBMS instance being a primary handler of requests to interact with the data store and the second DBMS instance being a shadow instance that monitors the first DBMS instance, the method comprising:

initiating, by the first DBMS instance, at least one first process associated with execution of a request to interact with the data store;

notifying, by the first DBMS instance using the communication facility, the second DBMS instance of the at least one first process;

initiating, by the second DBMS instance, at least one second process associated with the at least one first process; and

executing a command from the first DBMS instance to open a dataset in the data store, wherein notifying, by the first DBMS instance, the second DBMS instance comprises communicating a message from the first DBMS instance to the second DBMS instance, the message notifying the second DBMS instance of the dataset to be opened, and wherein the at least one second process comprises executing a second command from the second DBMS instance to open the dataset based, at least in part, on the message,

wherein the dataset to be opened is opened in parallel by the first DBMS instance and the second DBMS instance, whereby the dataset to be opened is opened by the second DBMS instance before failure of the first DBMS instance.

13. The method of claim 12, further comprising:

executing a data processing query from the first DBMS instance using the opened dataset; and

communicating information from the first DBMS instance associated with the data processing query to the communication facility.

14. The method of claim 13, further comprising:

generating a log by the first DBMS instance, the log comprising information related to data processing of the first DBMS instance; and

monitoring data processing of the first DBMS instance at the second DBMS instance using the communication facility to scan the log.

15. The method of claim 14, further comprising performing data processing using the second DBMS instance in response to a failure of the first DBMS instance.

16. The method of claim 15, wherein performing data processing using the second DBMS instance in response to a failure of the first DBMS instance comprises:

reprocessing a pipeline executed by the first DBMS instance on the data store using the second DBMS based, at least in part, on the scanned log; and

in response to identifying a rollback, scheduling the rollback for the data store using the second DBMS.

17. The method of claim 13, further comprising:

executing a command to close the dataset using the first DBMS instance;

communicating a message from the first DBMS instance to the second DBMS instance, the message notifying the second DBMS instance of the dataset to be closed; and

executing a second command from the second DBMS to close the dataset based, at least in part, on the message.

18. The method of claim 12, further comprising:

receiving, by the second DBMS instance, notification that the first DBMS instance failed to execute the request; and

attempting, by the second DBMS instance, to execute the request using, at least in part, the at least one second process.

19. The method of claim 12, the second DBMS instance comprising a local DBMS instance.

20. The method of claim 19 further comprising monitoring data processing of the first DBMS instance at a third DBMS instance using the communication facility communicably coupled to the data store, the third DBMS instance comprising a second shadow instance operable to and the third DBMS instance comprising a remote DBMS instance.

21. The method of claim 12, the second DBMS instance comprising a remote DBMS instance.

22. The method of claim 12, wherein executing the first DBMS instance comprises executing the first DBMS instance in a first logical partition and wherein executing the second DBMS comprises executing the remote second DBMS instance in a second logical partition.

23. A system for providing increased database fault tolerance in a data storage environment, wherein a first Database Management Service (DBMS) instance and a second DBMS instance are communicably coupled with a data store, and wherein the first DBMS instance is communicably coupled to the second DBMS instance through a communication facility, the first DBMS instance being a primary handler of requests to interact with the data store and the second DBMS instance being a shadow instance that monitors the first DBMS instance, the system comprising:

memory storing a data store; and

one or more processors that;

initiate, by the first DBMS instance, at least one first process associated with execution of a request to interact with the data store;

notify, by the first DBMS instance using the communication facility, the second DBMS instance of the at least one first process; and

initiate, by the second DBMS instance, at least one second process associated with the at least one first process,

wherein the at least one first process comprises execution of a command from the first DBMS instance to open a dataset in the data store,

wherein notification by the first DBMS instance to the second DBMS instance comprises communication of a message from the first DBMS instance to the second DBMS instance, the message notifying the second DBMS instance of the dataset to be opened, and wherein the at least one second process comprises execution of a second command from the second DBMS instance to open the dataset based, at least in part, on the message,

wherein the dataset to be opened is opened in parallel by the first DBMS instance and the second DBMS instance, whereby the dataset to be opened is opened by the second DBMS instance before failure of the first DBMS instance.

24. The system of claim 23, wherein the one or more processors further:

execute a data processing query from the first DBMS instance using the opened dataset; and

communicate information from the first DBMS instance associated with the data processing query to the communication facility.

25. The system of claim 24, wherein the one or more processors further:

generate a log by the first DBMS instance, the log comprising information related to data processing of the first DBMS instance; and

monitor data processing of the first DBMS instance at the second DBMS instance using the communication facility by instructing the second DBMS instance to scan the log to identify processing by the first DBMS instance.

26. The system of claim 25, wherein the one or more processors further automatically performs data processing using the second DBMS instance in response to a failure of the first DBMS instance.

27. The system of claim 26, wherein the one or more processors that perform data processing using the second DBMS instance in response to a failure of the first DBMS instance further:

reprocesses a pipeline executed by the first DBMS instance on the data store using the second DBMS based, at least in part, on the scanned log; and

in response to identifying a rollback, schedules the rollback for the data store using the second DBMS.

28. The system of claim 24, wherein the one or more processors further:

execute a command to close the dataset using the first DBMS instance;

communicate a message from the first DBMS instance to the second DBMS instance, the message notifying the second DBMS instance of the dataset to be closed; and

execute a second command from the second DBMS to close the dataset based, at least in part, on the message.

29. The system of claim 23, wherein the one or more processors further:

receive, by the second DBMS instance, notification that the first DBMS instance failed to execute the request; and

attempt, by the second DBMS instance, to execute the request using, at least in part, the at least one second process.

30. The system of claim 23, the second DBMS instance comprising a local DBMS instance.

31. The system of claim 30, wherein the one or more processors further monitor data processing of the first DBMS instance at a third DBMS instance using the communication facility communicably coupled to the data store, the third DBMS instance comprising a second shadow instance operable and comprising a remote DBMS instance.

32. The system of claim 23, the second DBMS instance comprising a remote DBMS instance.