Logical level predictive caching in relational databases
View Patent ↗A system, method, and computer-readable medium for performing a predictive caching operation in which a hit rate is improved by pre executing statements that are predicted based upon previous use patterns. More specifically, by analyzing a stream of statements provided to a server, such as a database server, patterns of usage are detected and based upon these patterns, statements provided to the server are executed before the actual application executes the predicted statements. Thus the application executes faster because the data requested is already cached based upon the predictive execution.
1. A computer-implementable method to perform a predictive caching operation, the method comprising:
analyzing a stream of statements provided to a database server;
determining a pattern of usage within the stream of statements;
determining, within a predetermined level of certainty, that the pattern of usage includes a first statement that will be executed after a second statement;
predictively pre-executing the first statement on the database server based on the pattern of usage within the stream of statements so as to improve performance of the server when executing the stream of statements, wherein predictively pre-executing the first statement comprises receiving rules relating to the stream of statements, each rule comprising a trigger statement and associated actions;
providing the pre-executed first statement for storage in a cache that is accessible to an application; and
upon detecting the first statement, retrieving the pre-executed first statement from the cache and initiating execution of the second statement.
2. The method of claim 1 , further comprising:
monitoring statements that are sent to the database server by an application; and,
collecting the statements that are sent to the database server.
3. The method of claim 2 , wherein:
the collecting further comprises collecting a plurality of characteristics relating to each statement, the plurality of characteristics comprising one or more of a query text, any optional bind variables and bind variable values for the query, a unique identifier for a session to the database and an amount of time spent executing the query.
4. The method of claim 1 , wherein:
the analyzing comprises processing the stream of statements and detecting patterns of execution within the stream of statements.
5. The method of claim 4 , wherein determining the patterns of usage further comprises performing a pattern matching operation, the pattern matching operation recognizing that an application obtains a detail record for a certain entity following several successive queries to obtain additional data for that entity.
6. A system comprising:
a processor;
a data bus coupled to the processor; and
a non-transitory, computer-readable storage medium embodying computer program code, the non-transitory, computer-readable storage medium being coupled to the data bus, the computer program code interacting with a plurality of computer operations and comprising instructions executable by the processor to perform operations comprising:
analyze a stream of statements provided to a database server;
determine a pattern of usage within the stream of statements;
determine, within a predetermined level of certainty, that the pattern of usage includes a first statement that will be executed after a second statement;
predictively pre-execute the first statement on the database server based on the pattern of usage within the stream of statements so as to improve performance of the server when executing the stream of statements, wherein predictively pre-executing the first statement comprises receiving rules relating to the stream of statements, each rule comprising a trigger statement and associated actions;
provide the pre-executed first statement for storage in a cache that is accessible to an application; and
upon detecting the first statement, retrieve the pre-executed first statement from the cache and initiate execution of the second statement.
7. The system of claim 6 , wherein the instructions are further configured to cause the processor to perform operations comprising:
monitor statements that are sent to the database server by an application; and;
collect statements that are sent to the database server.
8. The system of claim 7 , wherein execution of the instruction to collect causes the processor to:
collect a plurality of characteristics relating to each statement, the plurality of characteristics comprising one or more of a query text, any optional bind variables and bind variable values for the query, a unique identifier for a session to the database and an amount of time spent executing the query.
9. The system of claim 6 , wherein execution of the instruction to analyze causes the processor to:
process the stream of statements and detect patterns of execution within the stream of statements.
10. The system of claim 9 , wherein execution of the instruction to determine the pattern of usage causes the processor to:
perform a pattern matching operation, the pattern matching operation recognizing that an application obtains a detail record for a certain entity following several successive queries to obtain additional data for that entity.
11. A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured to cause a processor to perform operations comprising:
analyze a stream of statements provided to a database server;
determine a pattern of usage within the stream of statements;
determine, within a predetermined level of certainty, that the pattern of usage includes a first statement that will be executed after a second statement;
predictively pre-execute the first statement on the database server based on the pattern of usage within the stream of statements so as to improve performance of the server when executing the stream of statements, wherein predictively pre-executing the first statement comprises receiving rules relating to the stream of statements, each rule comprising a trigger statement and associated actions;
provide the pre-executed first statement for storage in a cache that is accessible to an application; and
upon detecting the first statement, retrieve the pre-executed first statement from the cache and initiating execution of the second statement.
12. The non-transitory, computer-readable storage medium of claim 11 , wherein the computer executable instructions are further configured to cause the processor to:
monitor statements that are sent to the database server by an application; and,
collect statements that are sent to the database server.
13. The non-transitory, computer-readable storage medium of claim 11 , wherein:
when collecting the statements that are sent to the database server, the processor is configured to collect a plurality of characteristics relating to each statement, the plurality of characteristics comprising one or more of a query text, any optional bind variables and bind variable values for the query, a unique identifier for a session to the database and an amount of time spent executing the query.
14. The non-transitory, computer-readable storage medium of claim 11 , wherein:
when analyzing the stream of statements provided to the database server, the processor is configured to process the stream of statements and detect patterns of execution within the stream of statements.
15. The non-transitory, computer-readable storage medium of claim 14 , wherein:
when determining the pattern of usage, the processor is configured to perform a pattern matching operation, the pattern matching operation recognizing that an application obtains a detail record for a certain entity following several successive queries to obtain additional data for that entity.