State initialization for continuous queries over archived views

• US 9,805,095 B2
• Filed: 09/25/2013
• Issued: 10/31/2017
• Est. Priority Date: 09/28/2012
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Deshmukh, Unmesh Anil, Srinivasan, Anand, Shukla, Vikram
• Primary Examiner(s): Thomas, Ashish K
• Assistant Examiner(s): Ferrer, Jedidiah
• Application Number:
  US14/036,500
• Publication Number:
  US 20140095444A1
• Time in Patent Office: 1,497 Days
• CPC Class Codes: G06F 16/2308 Concurrency control transaction processing G06F9/466
• CPC Class Codes: G06F 16/24542 Plan optimisation
• CPC Class Codes: G06F 16/24544 Join order optimisation
• CPC Class Codes: G06F 16/24568 Data stream processing; Continuous queries
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A method may include receiving a continuous query identified by an archived view. The method may also include creating the archived view, the archived view identified by a join query related to two or more archived relations associated with an application and/or generating a query plan for the continuous query. The method may also include identifying a view root operator in the archived view, the view root operator identified as a join operator in the archived view and/or identifying a query operator in the query plan that topologically precedes the view root operator. Further, the method may include constructing an archiver query for the identified query operator that topologically precedes the view root operator, executing the archiver query to obtain a result set of data records related to the application, and/or generating a snapshot output of one or more data values related to the application.

Claims

1. A computer-implemented method, comprising:

receiving, by a computer system configured to execute computer-executable instructions, a continuous query, the continuous query identifying an archived view, the archived view identifying a join query between at least a first archived relation and a second archived relation associated with an application, the first archived relation comprising at least a first portion of historical data and incoming events related to an event stream of the application and the second archived relation comprising at least a second portion of the historical data and the incoming events related to the event stream;

generating a query plan for the continuous query;

identifying a view root operator in the archived view, the view root operator being identified as a join operator in the join query identified in the archived view;

determining that a set of one or more operators in the query plan comprises one or more query operators, the one or more query operators being identified as stateful query operators of the query plan that maintain state information related to the application;

identifying one or more of the stateful query operators in the query plan that topologically precede the view root operator;

constructing a set of one or more archiver queries corresponding to the one or more stateful query operators that topologically precede the view root operator;

combining the set of one or more archiver queries to generate a combined archiver query;

executing the combined archiver query to obtain a result set of data records related to the application;

converting the result set of data records into a list of tuples for each of the one or more stateful query operators by;

determining a value of an ordering attribute in the combined archiver query, the value of the ordering attribute corresponding to a number of the one or more stateful query operators in the query plan; and

constructing the list of tuples for each of the one or more stateful query operators based at least in part on the value of the ordering attribute;

initializing a state of the one or more stateful query operators based at least in part on the list of tuples; and

generating a snapshot output of one or more data values related to the application based at least in part on the state of the one or more stateful query operators.

2. The computer-implemented method of claim 1, further comprising:

traversing the query plan in topological order to identify the view root operator, the view root operator being identified by an output operator to store event information related to the application;

inserting a buffer operator between the view root operator and the output operator in the query plan; and

storing the event information from the view root operator in the buffer operator.

3. The computer-implemented method of claim 1, further comprising:

identifying a second query operator in a second continuous query concurrently executing as being the same type as the one or more query operators identified in the query plan, the second continuous query being identified based at least in part on the archived view; and

generating a combined query plan based at least in part on the one or more query operators identified in the query plan being the same type as the second query operator in the second continuous query.

4. The computer-implemented method of claim 1, further comprising initializing the state of the one or more stateful query operators by constructing a list of tuples for the one or more stateful query operators from the result set of data records, wherein at least one tuple in the list of tuples is constructed based at least in part on a subset of attributes in the result set of data records.

5. The computer-implemented method of claim 1, further comprising displaying the snapshot output of the one or more data values related to the application to a user of the application via a display device.

6. The computer-implemented method of claim 1, further comprising processing incoming real-time events related to the application based at least in part on the snapshot output of the one or more data values.

7. A non-transitory computer-readable storage medium storing computer-executable instructions that, when executed by one or more computer systems, configure the one or more computer systems to perform at least:

instructions that cause the one or more computer systems to receive a continuous query, the continuous query identifying an archived view, the archived view identifying a join query between-at least a first archived relation and a second archived relation associated with an application, the first archived relation comprising at least a first portion of historical data and incoming events related to an event stream of the application and the second archived relation comprising at least a second portion of the historical data and the incoming events related to the event stream;

instructions that cause the one or more computer systems to identify a view root operator in the archived view, the view root operator being identified as a join operator in the join query identified in the archived view;

instructions that cause the one or more processors to determine that a set of one or more operators in a query plan for the continuous query comprises one or more query operators, the one or more query operators being identified as stateful query operators of the query plan that maintain state information related to the application;

instructions that cause the one or more computer systems to identify one or more of the stateful query operators in the query plan that topologically precede the view root operator;

instructions that cause the one or more computer systems to construct a set of one or more archiver queries corresponding to the one or more stateful query operators that topologically precede the view root operator;

instructions that cause the one or more computer systems to combine the set of one or more archiver queries to generate a combined archiver query;

instructions that cause the one or more computer systems to execute the combined archiver query to obtain a result set of data records related to the application;

instructions that cause the one or more computer systems to convert the result set of data records into a list of tuples for each of the one or more stateful query operators by;

determining a value of an ordering attribute in the combined archiver query, the value of the ordering attribute corresponding to a number of the one or more stateful query operators in the query plan; and

constructing the list of tuples for each of the one or more stateful query operators based at least in part on the value of the ordering attribute;

instructions that cause the one or more computer systems to initialize a state of the one or more stateful query operators based at least in part on the list of tuples; and

instructions that cause the one or more computer systems to generate a snapshot output of one or more data values related to the application based at least in part on the state of the one or more stateful query operators.

8. The non-transitory computer-readable storage medium of claim 7, further comprising instructions that cause the one or more computer systems to initialize the state of the one or more stateful query operators by constructing a list of tuples for the one or more identified query operators from the result set of data records.

9. The non-transitory computer-readable storage medium of claim 8, wherein at least one tuple in the list of tuples is constructed based at least in part on a subset of attributes in the result set of data records.

10. The non-transitory computer-readable storage medium of claim 7, further comprising instructions that cause the one or more computer systems to at least display a snapshot output of one or more data values related to the application based at least in part on the result set of data records related to the application.

11. A system, comprising:

a memory storing a plurality of instructions; and

a central processing unit configured to access the memory, wherein the central processing unit is further configured to execute the plurality of instructions to at least;

receive a continuous query, the continuous query identifying an archived view, the archived view identifying a join query between-at least a first archived relation and a second archived relation associated with an application, the first archived relation comprising at least a first portion of historical data and incoming events related to an event stream of the application and the second archived relation comprising at least a second portion of the historical data and the incoming events related to the event stream;

generate a query plan for the continuous query;

identify a view root operator in the archived view, the view root operator being identified as a join operator in the join query identified in the archived view;

determine that a set of one or more operators in the query plan comprises one or more query operators, the one or more query operators being identified as stateful query operators of the query plan maintain state information related to the application;

identify one or more of the stateful query operators in the query plan that topologically precede the view root operator;

construct a set of one or more archiver queries corresponding to the one or more stateful query operators that topologically precede the view root operator;

combine the set of one or more archiver queries to generate a combined archiver query;

execute the combined archiver query to obtain a result set of data records related to the application;

convert the result set of data records into a list of tuples for each of the one or more stateful query operators by;

determining a value of an ordering attribute in the combined archiver query, the value of the ordering attribute corresponding to a number of the one or more stateful query operators in the query plan; and

constructing the list of tuples for each of the one or more stateful query operators based at least in part on the value of the ordering attribute;

initialize a state of the one or more stateful query operators based at least in part on the list of tuples; and

generate a snapshot output of one or more data values related to the application based at least in part on the state of the one or more stateful query operators.

12. The system of claim 11, wherein the central processing unit is further configured to execute the plurality of instructions to at least:

traverse the query plan in topological order to identify the view root operator, the view root operator being identified by an output operator configured to store event information related to the application;

insert a buffer operator between the view root operator and the output operator in the query plan; and

store the event information from the view root operator in the buffer operator.

13. The system of claim 11, wherein the central processing unit is further configured to execute the plurality of instructions to at least:

identify a second query operator in a continuous query concurrently executing as being the same type as the one or more query operators in the query plan; and

generate a combined query plan based at least in part on the one or more query operators identified in the query plan being the same type as the second query operator in the continuous query.