Support for user defined aggregations in a data stream management system

• US 8,204,875 B2
• Filed: 07/16/2011
• Issued: 06/19/2012
• Est. Priority Date: 10/20/2007
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Srinivasan, Anand, Jain, Namit, Mishra, Shailendra Kumar
• Primary Examiner(s): Vy, Hung T
• Application Number:
  US13/184,528
• Publication Number:
  US 20110270879A1
• Time in Patent Office: 339 Days
• Field of Search: 707/713, 707/714, 707/715, 707/718, 707/769
• US Class Current: 707/713
• CPC Class Codes: G06F 16/24556 Aggregation; Duplicate elimination
• CPC Class Codes: G06F 16/24568 Data stream processing; Continuous queries

Abstract

A computer is programmed to accept a command to create a new aggregation defined by a user during execution of continuous queries on streams of data. The computer is further programmed to thereafter accept and process new continuous queries using the new aggregation, in a manner similar to built-in aggregations. The user typically writes a set of instructions to perform the new aggregation, and identifies in the command, a location of the set of instructions. In response to such a command, the computer creates metadata identifying the new aggregation. The metadata is used to instantiate one aggregation for each group of data in a current window, grouped by an attribute identified in a new query.

Claims

1. A data stream management system that processes a plurality of streams of data using a plurality of continuous queries, the data stream management system comprising:

a store comprising a plurality of tuples representing a relation, each tuple being time stamped;

a memory comprising a global plan currently being used in execution of the continuous queries on the plurality of streams;

a processor, coupled to said store, to modify said global plan in said memory in response to receipt of a new continuous query that uses a new aggregation defined by a command, by using an operator to invoke a first instance in a plurality of instances of the new aggregation based on a first value of an attribute of said data;

wherein said new aggregation groups data in said streams by said attribute;

wherein the processor is configured to execute an instruction in the operator, to perform the new aggregation;

wherein a first count corresponds to said first value of said attribute and said first count is changed for data of said first value by the processor;

wherein the processor is configured to create the first instance when the first count is changed from an initial value; and

wherein the processor is configured to delete the first instance when the first count is changed to return to said initial value.

2. The data stream management system of claim 1 wherein:

said operator comprises a plurality of additional counts for a corresponding plurality of additional values of said attribute.

3. The data stream management system of claim 1 wherein:

the processor is configured to check to ensure said new aggregation is identified in said new continuous query in a select list.

4. The data stream management system of claim 1 wherein:

the operator comprises a plurality of structures to hold a corresponding plurality of paths to said plurality of instances of the new aggregation.

5. The data stream management system of claim 1 wherein:

each instance in said plurality of instances of the new aggregation holds state information specific to a corresponding value of the attribute.

6. The data stream management system of claim 1 wherein:

the processor is configured to create and delete each instance in said plurality of instances in a just-in-time manner, when a window comprising data to be processed by said each instance becomes non-empty and empty respectively.

7. One or more non-transitory computer-readable storage media comprising software including instructions to process a plurality of streams of data, the instructions comprising:

instructions to store in a store a plurality of tuples representing a relation, each tuple being time stamped;

instructions to use a global plan in a memory to execute a plurality of continuous queries on the plurality of streams;

instructions responsive to receipt of a new continuous query that uses a new aggregation defined by a user command, to modify in said memory said global plan currently being used in execution of the continuous queries, by using an operator to invoke a first instance in a plurality of instances of the new aggregation based on a first value of an attribute of said data;

wherein on execution of said instructions to modify, said new aggregation groups data in said streams by said attribute;

instructions to perform the new aggregation when using said operator;

wherein a first count corresponds to said first value of said attribute and said first count is changed for data of said first value by execution of the instructions to perform;

instructions to create the first instance when the first count is changed from an initial value; and

instructions to delete the first instance when the first count is changed to return to said initial value.

8. The one or more non-transitory computer-readable storage media of claim 7 wherein:

said instructions to perform operate only on data of said first value of said attribute.

9. The one or more non-transitory computer-readable storage media of claim 7 wherein:

said operator comprises a plurality of additional counts for a corresponding plurality of additional values of said attribute.

10. The one or more non-transitory computer-readable storage media of claim 7 further comprising:

instructions to check to ensure said new aggregation is identified in said new continuous query in a select list.

11. The one or more non-transitory computer-readable storage media of claim 7 wherein:

the operator comprises a plurality of structures to hold a corresponding plurality of paths to said plurality of instances of the new aggregation.

12. The one or more non-transitory computer-readable storage media of claim 7 wherein:

each instance in said plurality of instances of the new aggregation holds state information specific to a corresponding value of the attribute.

13. The one or more non-transitory computer-readable storage media of claim 7 further comprising:

instructions to create and delete each instance in said plurality of instances in a just-in-time manner, when a window comprising data to be processed by said each instance becomes non-empty and empty respectively.

14. A method implemented in a computer of processing a plurality of streams of data, the method comprising:

storing in a store a plurality of tuples representing a relation, each tuple being time stamped;

using a global plan in a memory to execute a plurality of continuous queries on the plurality of streams;

in response to receipt of a new continuous query that uses a new aggregation defined by a command, modifying in said memory said global plan currently being used in execution of the continuous queries, by using an operator to invoke a first instance in a plurality of instances of the new aggregation based on a first value of an attribute of said data;

wherein said new aggregation groups data in said streams by said attribute;

executing at least one instruction in the operator, to perform the new aggregation;

wherein a first count corresponds to said first value of said attribute and said first count is changed for data of said first value by said executing;

creating the first instance when the first count is changed from an initial value; and

deleting the first instance when the first count is changed to return to said initial value.

15. The method of claim 14 wherein:

said executing of at least one instruction operates only on data of said first value of said attribute.

16. The method of claim 14 wherein:

said operator comprises a plurality of additional counts for a corresponding plurality of additional values of said attribute.

17. The method of claim 14 further comprising:

checking to ensure said new aggregation is identified in said new continuous query in a select list.

18. The method of claim 14 wherein:

the operator comprises a plurality of structures to hold a corresponding plurality of paths to said plurality of instances of the new aggregation.

19. The method of claim 14 wherein:

each instance in said plurality of instances of the new aggregation holds state information specific to a corresponding value of the attribute.

20. The method of claim 14 further comprising:

creating and deleting each instance in said plurality of instances in a just-in-time manner, when a window comprising data to be processed by said each instance becomes non-empty and empty respectively.