Addition and processing of continuous SQL queries in a streaming relational database management system

• US 8,745,070 B2
• Filed: 01/05/2012
• Issued: 06/03/2014
• Est. Priority Date: 03/06/2008
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Krishnamurthy, Saileshwar, Thombre, Neil, Conway, Neil, Li, Wing Hang, Hoyer, Morten
• Primary Examiner(s): Btit, Jacob F.
• Assistant Examiner(s): KIM, CHRISTY Y
• Application Number:
  US13/344,559
• Publication Number:
  US 20120124096A1
• Time in Patent Office: 880 Days
• Field of Search: 707/751
• US Class Current: 707/751
• CPC Class Codes: G06F 16/24532 of parallel queries
• CPC Class Codes: G06F 16/24568 Data stream processing; Continuous queries
• CPC Class Codes: G06F 16/2471 Distributed queries

Abstract

Systems, methods, and media are disclosed herein that can be embodied in a traditional Relational Database Management System (RDBMS) in order to transform it into a Streaming Relational Database Management System (SRDBMS). An SRDBMS may provide functionality such as to manage and populate streams, tables, and archived stream histories and support the evaluation of continuous queries on streams and tables. Both continuous and snapshot queries support the full spectrum of the industry standard, widely used, Structured Query Language. The present technology can support a high number of concurrent continuous queries using a scalable and efficient shared query evaluation scheme, support on-the-fly addition of continuous queries into a mechanism that implements the shared evaluation scheme, reuse RDBMS modules such as relational operators and expression evaluators, and visualize results of continuous queries in real time.

Claims

1. A method for managing a data stream, comprising:

a streaming relational database management server receiving a continuous query, wherein the continuous query is a request for an unbounded set of records over time;

in response to receiving the continuous query, the streaming relational database management server generating a first stream-to-relation operator;

the first stream-to-relation operator returning, to a second stream-to-relation operator, an unbounded set of windows over time, wherein a first window comprises a tuple comprising;

a first control information that indicates a first edge, the first edge belonging to the first window;

a second control information that indicates a second edge, the second edge belonging to a second window;

using the second stream-to-relation operator, determining the first edge for the first window based on the first control information;

using the second stream-to-relation operator, determining the second edge for the second window based on the second control information;

based, at least in part, on the second control information, the second stream-to-relation operator returning a window before receiving the second window;

wherein the method is performed by one or more computing devices.

2. The method of claim 1, further comprising:

the first stream-to-relation operator returning the unbounded set of windows to the streaming relational database management server;

in response to the streaming relational database management server receiving the first window of the unbounded set of windows from the first stream-to-relation operator with at least one record, returning the at least one record within the first window to a client.

3. The method of claim 1, wherein the first control information is a row number, and the first stream-to-relation operator returning the first window of the unbounded set of windows based on the row number.

4. The method of claim 1, wherein the first control information is a timestamp, and the first stream-to-relation operator returning the first window of the unbounded set of windows based on the timestamp.

5. The method of claim 1, wherein the first window comprises a data tuple, which comprises a record from the unbounded set of records, and the first control information is an attribute of the record in the data tuple.

6. The method of claim 2, comprising displaying a result of one or more records of the at least one record returned by the client.

7. One or more non-transitory computer-readable media storing instructions which, when executed, cause:

a streaming relational database management server receiving a continuous query, wherein the continuous query is a request for an unbounded set of records over time;

in response to receiving the continuous query, the streaming relational database management server generating a first stream-to-relation operator;

the first stream-to-relation operator returning, to a second stream-to-relation operator, an unbounded set of windows over time, wherein a first window comprises tuples a tuple comprises;

a first control information that indicates a first edge, the first edge belonging to the first window;

a second control information that indicates a second edge, the second edge belonging to a second window;

using the second stream-to-relation operator, determining the first edge for the first window based on the first control information;

using the second stream-to-relation operator, determining the second edge for the second window based on the second control information;

based, at least in part, on the second control information, the second stream-to-relation operator returning a window before receiving the second window.

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

the first stream-to-relation operator returning the unbounded set of windows to the streaming relational database management server;

in response to the streaming relational database management server receiving the first window of the unbounded set of windows from the first stream-to-relation operator with at least one record, returning the at least one record within the first window to a client.

9. The one or more non-transitory computer-readable media of claim 8, comprising displaying a result of one or more records of the at least one record returned by the client.

10. The one or more non-transitory computer-readable media of claim 7, wherein the first control information is a timestamp, and the first stream-to-relation operator returning the first window of the unbounded set of windows based on the timestamp.

11. The one or more non-transitory computer-readable media of claim 7, wherein the first control information is a row number, and the first stream-to-relation operator returning the first window of the unbounded set of windows based on the row number.

12. The one or more non-transitory computer-readable media of claim 7, wherein the first window comprises a data tuple, which comprises a record from the unbounded set of records, and the first control information is an attribute of the record in the data tuple.

13. One or more non-transitory computer-readable media storing instructions which, when executed, cause:

a streaming relational database management server receiving a continuous query, wherein the continuous query is a request for an unbounded set of records over time, and includes a subquery and a superquery;

in response to receiving the continuous query the streaming relational database management server generating a first stream-to-relation operator associated with the subquery and a second stream-to-relation operator associated with the superquery;

the first stream-to-relation operator returning an unbounded first set of windows over time to the second stream-to-relation operator;

wherein a window from the unbound first set of windows comprises one or more control tuples, and a control tuple comprises;

a first control information that is used to determine a first window edge;

a second control information, wherein the second control information may be used for further downstream window processing;

in response to the second stream-to-relation operator receiving one or more windows of the unbound first set of windows from the first stream-to-relation operator, the second stream-to-relation operator returning an unbound second set of windows over time;

wherein a window from the unbound second set of windows comprises one or more control tuples, from the one or more windows received from the first stream-to-relation operator, based on the second control information stored in each of the one or more control tuples.

14. The one or more non-transitory computer-readable media of claim 13, wherein each window in the unbounded first set of windows spans a shorter amount of time than each window in the unbounded second set of windows.

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

the second stream-to-relation operator returning one or more windows of the unbounded second set of windows to the streaming relational database management server;

in response to the streaming relational database management server receiving the one or more windows from the second stream-to-relation operator, returning the records within the received one or more windows, from the second stream-to-relation operator, to a client.

16. The one or more non-transitory computer-readable media of claim 15, comprising displaying a result of one or more records of the records returned to the client.

17. The one or more non-transitory computer-readable media of claim 13:

wherein the first control information is a timestamp and the first stream-to-relation operator returning one or more windows of the unbounded first set of windows based on the timestamp;

wherein the second control information is a speculative timestamp of a subsequent window returned by the first stream-to-relation operator.

18. The one or more non-transitory computer-readable media of claim 13:

wherein the first control information is a row number and the first stream-to-relation operator returning one or more windows of the unbounded first set of windows based on the row number;

wherein the second control information is a speculative row number of a subsequent window returned by the first stream-to-relation operator.

19. The one or more non-transitory computer-readable media of claim 13, wherein the window comprises a data tuple, which comprises a record from the unbounded set of records, and the first control information is an attribute of the record in the data tuple.

20. A method for managing a data stream, comprising:

a streaming relational database management server receiving a continuous query, wherein the continuous query is a request for an unbounded set of records over time, and includes a subquery and a superquery;

in response to receiving the continuous query, the streaming relational database management server generating a first stream-to-relation operator associated with the subquery, and a second stream-to-relation operator associated with the superquery;

the first stream-to-relation operator returning an unbounded first set of windows over time to the second stream-to-relation operator;

in response to the second stream-to-operator receiving one or more windows of the unbounded first set of windows, the second stream-to-relation operator returning an unbounded second set of windows over time;

wherein a first window of the unbounded first set of windows comprises a control tuple comprising;

a first control information;

a second control information;

in response to the second stream-to-relation operator receiving the first window, the second stream-to-relation operator;

determining a first leading window edge belonging to the first window based on the first control information;

determining a second leading window edge belonging to a second window from the unbounded first set of windows based on the second control information;

determining that a window, from the unbounded second set of windows, has a trailing window edge that is less than the second leading window edge;

in response to determining that the trailing window edge is less than the second leading window edge, returning the window;

wherein the method is performed by one or more computing devices.