TECHNIQUES FOR REAL-TIME DATABASE PROCESSING

• US 20100153351A1
• Filed: 12/16/2008
• Published: 06/17/2010
• Est. Priority Date: 12/16/2008
• Status: Abandoned Application

• Current Assignee: Teradata Corporation
• Original Assignee: Teradata Corporation
• Inventors: Burlew, Thomas William, Yung, Alex P.
• Application Number:
  US12/335,853
• Publication Number:
  US 20100153351A1
• Time in Patent Office: Days
• US Class Current: 707/703
• CPC Class Codes: G06F 11/1458 Management of the backup or restore process
• CPC Class Codes: G06F 16/2379 Updates performed during online database operations; commit processing
• CPC Class Codes: G06F 2201/80 Database-specific techniques

Abstract

Techniques for real time data processing are provided. Unprocessed database transactions are represented in messages that are captured in files and stored in directories. The order of the files can be user-defined within the directory. The files are then processed to committed transactions that are committed to a database for updating. Checkpoints can occur at user-defined or data-driven points for failover recovery processing.

Claims

1. A method implemented in a computer-readable storage medium and processed by a processor to perform the method, comprising:

scanning a directory for database transactions defined in files located within the directory;

processing each of the files in a user-defined order as the database transaction are applied to a database; and

initiating a checkpoint operation to indicate each of the files have been processed to the database.

2. The method of claim 1 further comprising, iterating the method processing at a user-defined start time and at user-defined intervals until a user-defined end time is reached.

3. The method of claim 1 further comprising, iterating the method processing 24 hours a day and 7 days a week for 365 days a year at user-defined intervals.

4. The method of claim 1, wherein scanning further includes reading multiple different files in parallel with one another, each different file having a different schema definition from the others.

5. The method of claim 1, wherein scanning further includes initiating user-defined or data-driven checkpoint operations during the scanning process.

6. The method of claim 1, wherein processing further includes recognizing the user-defined order as a time sequence, wherein each of the files are named within the directory having date and time stamps as part of their names to assist in resolving the time sequence for the user-defined order.

7. The method of claim 1, wherein initiating further includes moving each of the files once the checkpoint operation successfully completes to an archive directory.

8. A method implemented in a computer-readable storage medium and processed by a processor to perform the method, comprising:

moving unprocessed database transactions defined as messages to files that are housed in one or more directories on one or more servers;

detecting the files in the one or more directories and processing the unprocessed database transactions parsed from the files in a user-defined order to a database; and

taking one or more checkpoints for failover support should processing fail at some point in time while processing the unprocessed database transactions.

9. The method of claim 8, wherein moving further includes including multiple dependent messages in a single file.

10. The method of claim 8, wherein moving further includes creating each file with a user-defined name that includes as least a portion of that user-defined name a date and time stamp associated with a corresponding message for when that corresponding message was generated.

11. The method of claim 8, wherein detecting further includes reading multiple files in parallel, wherein at least two of the files read in parallel include different database formats for their corresponding embedded messages.

12. The method of claim 8, wherein detecting further includes moving each file out of the one or more directories to an archive directory once each unprocessed database transaction associated with that file is processed.

13. The method of claim 8, wherein taking further includes performing at least one checkpoint when a single file or a predefined number of files are successfully processed to the database.

14. The method of claim 8, taking further includes performing at least one checkpoint that is user-defined and at least another checkpoint that is data driven and defined by a policy or event that is evaluated or detected.

15. A computer-implemented system that executes on one or more processors, the system comprising:

a file generator implemented in a computer-readable storage medium and to execute on a processor of a network; and

a database transaction commit service implemented in a computer-readable storage medium and to execute on the processor or a different processor of the network;

wherein the file generator gathers messages for unprocessed database transactions from a plurality of disparate sources and produces files having these unprocessed database transactions, and wherein the file generator stores the files in one or more directories on one or more servers, and wherein the database transaction commit service periodically scans the one or more directories for the files and commits the embedded unprocessed database transactions to a database for processing.

16. The system of claim 15, wherein the file generator names the files within the one or more directories with user-defined naming criteria.

17. The system of claim 15, wherein the file generator sorts the files within the one or more directories via user-defined sort criteria.

18. The system of claim 15, wherein the database transaction commit service processes each file as soon as it is detected as being placed in the one or more directories.

19. The system of claim 15, wherein the database transaction commit service scans the one or more directories for the files at user-defined intervals.

20. The system of claim 15, wherein the database transaction commit service takes checkpoints each time a particular file and its unprocessed database transaction has been successfully committed to the database for processing.