Management of searches in a database system

• US 9,229,968 B2
• Filed: 03/11/2013
• Issued: 01/05/2016
• Est. Priority Date: 03/11/2013
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bestgen, Robert J., Beuch, Daniel E.
• Primary Examiner(s): PYO, MONICA M
• Application Number:
  US13/793,451
• Publication Number:
  US 20140258216A1
• Time in Patent Office: 1,030 Days
• Field of Search: 707/609, 707/610, 707/618, 707/696
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2272 Management thereof
• CPC Class Codes: G06F 16/2453 Query optimisation

Abstract

A method, system, and computer program product to efficiently search a database is disclosed. The method, system, and computer program product may include structuring the database to have a table, an index associated with the table, and a log associated with the index, the log adapted to include an entry. The method, system, and computer program product may include receiving a search request related to the index. The method, system, and computer program product may include searching the index when the search request is not associated with the entry. The method, system, and computer program product may include accounting for the entry and searching the index when the search request is associated with the entry.

Claims

1. A computer-implemented method of managing a database contained in a storage facility, comprising:

structuring the database to have a table, an index associated with the table, and a log associated with the index, wherein the log is configured to include an entry;

receiving a first search request related to the index;

determining that the first search request is not associated with the entry, wherein;

the first search request is associated with the entry if the entry affects any information sought by the first search request, wherein the entry affects information sought by the first search request if a pipeline between an application and the index is open and active, and

the first search request is not associated with the entry if;

a pipeline between an application and the index is either closed or open, but inactive, wherein the pipeline being either closed or open, but inactive includes;

the pipeline being open, but inactive, wherein the pipeline being open, but inactive includes;

receiving a second search request, the second search request being associated with the application, and wherein the index that is connected by the pipeline, which is open, is not actively in use;

the pipeline being either closed or open, but inactive includes the pipeline being closed, wherein the pipeline being closed includes a determination that the pipeline connecting the application and the index is not associated with a previous search request; and

searching the index in response to determining that the search request is not associated with the entry.

2. The method of claim 1, wherein the search request is not associated with the entry if the log contains zero entries.

3. The method of claim 1, further comprising:

determining that the search request is associated with the entry;

accounting for the entry in response to determining that the search request is associated with the entry, wherein accounting for the entry includes updating at least a portion of the index; and

searching the index using the search request.

4. The method of claim 3, wherein updating at least a portion of the index includes:

pausing, in response to determining the search request is associated with the entry, search of the index;

changing, in response to pausing search of the index, a row of the index; and

resuming, in response to changing the row of the index, search of the index.

5. The method of claim 1, further comprising:

determining the search request is associated with the entry;

searching the index using the search request; and

accounting for the entry in response to determining that the search request is associated with the entry, wherein accounting for the entry includes searching the log and merging results from searching the log and searching the index.

6. The method of claim 5, wherein searching the log and merging results from searching the log and searching the index includes:

determining the log has a number of entries below a threshold;

searching, using the search request, the log in response to determining the log has the number of entries below the threshold;

merging, in response to both searching the index and searching the log, results of both searching the index and searching the log; and

generating a final answer for the search request based on the merged results.

7. The method of claim 1, further comprising:

determining that the search request is associated with the entry;

searching the index using the search request; and

accounting for the entry in response to determining the search request is associated with the entry, wherein accounting for the entry includes answering with a result associated with the entry.

8. The method of claim 1, wherein the search request is not associated with the entry when the entry does not affect information sought by the search request.

9. The method of claim 8, wherein the entry does not affect information sought by the search request when a first data value of the entry does not meet a first data value threshold of the search request.

10. The method of claim 9, wherein the entry affects information sought by the search request when a second data value of the entry meets a second data value threshold of the search request.

11. The method of claim 1, wherein the determining that the pipeline is not both open and active includes a determination of a lack of data flow in the pipeline.

12. The method of claim 1, further comprising;

batching updates, in response to determining the search request is not associated with the entry, wherein batching updates includes configuring an update of multiple rows for one operation for performance at a later time to reduce input/output costs, working set memory sizes, central processing unit consumption, and performance overhead.

13. A computer-implemented method of managing a database contained in a storage facility, comprising:

structuring the database to have a table, an index associated with the table, and a log associated with the index, the log configured to include an entry;

receiving a search request related to the index;

determining, using both the entry and information sought by the search request, that the search request is not associated with the entry, wherein;

the first search request is associated with the entry if the entry affects any information sought by the first search request, wherein the entry affects information sought by the first search request if a pipeline between an application and the index is open and active, and

the first search request is not associated with the entry if;

a pipeline between an application and the index is either closed or open, but inactive, wherein the pipeline being either closed or open, but inactive includes;

the pipeline being open, but inactive, wherein the pipeline being open, but inactive includes;

receiving a second search request, the second search request being associated with the application, and wherein the index that is connected by the pipeline, which is open, is not actively in use;

the pipeline being either closed or open, but inactive includes the pipeline being closed, wherein the pipeline being closed includes a determination that the pipeline connecting the application and the index is not associated with a previous search request; and

searching the index without accounting for the entry in response to determining that the search request is not associated with the entry; and

accounting for the entry and searching the index in response to determining the search request is not associated with the entry.