System and method for writing data dependent upon multiple reads in a distributed database database

• US 20090210428A1
• Filed: 02/19/2008
• Published: 08/20/2009
• Est. Priority Date: 02/19/2008
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Srivastava, Utkarsh, Cooper, Brian, Yerneni, Ramana V.
• Granted Patent: US 7,895,172 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/273 Asynchronous replication or reconciliation

Abstract

An improved system and method for writing data dependent upon multiple reads in a distributed database is provided. A client may read several data records and may then send a request to a database server to perform a transaction to write a data record dependent upon multiple data records read. A database server may receive the request specifying a transaction to write a data record dependent upon multiple data records read and may perform the transaction by latching a master data record to be written and validating the data records the write depends upon. The multiple data records upon which the write depends may be validated by verifying the multiple data records are current versions of the data records stored in the distributed database. Data intensive applications may use this transaction type in large scale distributed database systems to provide stronger consistency without significantly degrading performance and scalability.

Claims

1. A distributed computer system for updating data tables, comprising:

a plurality of database servers operably coupled to provide a distributed database stored across the plurality of database servers;

a query processor operably coupled to at least one of the plurality of database servers for performing a transaction to update data by writing a single data record dependent upon a plurality of data records previously read from the distributed database; and

a storage manager operably coupled to the query processor for writing the single data record to a data table in the distributed database.

2. The system of claim 1 further comprising a plurality of client computers operably coupled to the plurality of database servers for sending a request to performing the transaction to update data by writing the single data record dependent upon the plurality of data records previously read from the distributed database.

3. A computer-readable medium having computer-executable components comprising the system of claim 1.

4. A computer-implemented method for updating data tables, comprising:

sending a plurality of requests to read data records in a distributed database;

receiving a plurality of data records read from a distributed database;

sending a request to perform a transaction to write a single data record dependent upon a plurality of data records read from the distributed database; and

receiving an acknowledgment in response to receiving the request to perform the transaction to write the single data record dependent upon the plurality of data records read from the distributed database.

5. The method of claim 4 further comprising receiving the request to perform the transaction to write the single data record dependent upon the plurality of data records read from the distributed database.

6. The method of claim 4 further comprising obtaining a local lock on the single data record in the distributed database.

7. The method of claim 4 further comprising validating that the plurality of data records received are current versions of the plurality of data records in the distributed database.

8. The method of claim 7 further comprising:

determining that the plurality of data records received are not current versions of the plurality of data records; and

sending a message indicating that the transaction failed to write the single data record dependent upon the plurality of data records read from the distributed database.

9. The method of claim 7 further comprising:

determining that the plurality of data records received are current versions of the plurality of data records; and

writing the single data record to a data table.

10. The method of claim 9 further comprising releasing the local lock on the single data record.

11. The method of claim 9 further comprising sending a message indicating that the transaction to write the single data record dependent upon the plurality of data records read from the distributed database succeeded.

12. The method of claim 7 wherein validating that the plurality of data records received are current versions of the plurality of data records comprises:

reading the current versions of the plurality of data records from the distributed database; and

determining whether there is any local lock on any of the current versions of the plurality of data records.

13. The method of claim 12 further comprising determining whether each of the plurality of data records received match the current versions of the plurality of data records.

14. The method of claim 12 further comprising:

determining that there is a local lock on at least one of the current version of the plurality of data records; and

returning an indication that the multiple data records read are invalid.

15. The method of claim 13 further comprising:

determining that there is not a local lock on any of the current version of the plurality of data records;

determining that each of the plurality of data records received match the current versions of the plurality of data records; and

returning an indication that the multiple data records read are valid.

16. The method of claim 13 further comprising:

determining that there is not a local lock on any of the current version of the plurality of data records;

determining that at least one of the plurality of data records received does not match the current version of the data record; and

returning an indication that the multiple data records read are invalid.

17. A computer-readable medium having computer-executable instructions for performing the method of claim 4.

18. A distributed computer system for updating data tables, comprising:

means for receiving a request to perform a transaction to write a single data record dependent upon a plurality of data records read from a distributed database;

means for obtaining a local lock on the single data record in the distributed database;

means for validating that the plurality of data records are current versions of the plurality of data records in the distributed database; and

means for sending an acknowledgment in response to receiving the request to perform the transaction to write the single data record dependent upon the plurality of data records read from the distributed database.

19. The computer system of claim 18 further comprising:

means for writing the single data record to a data table in the distributed database; and

means for releasing the local lock on the single data record in the distributed database.

20. The computer system of claim 18 further comprising:

means for sending a plurality of requests to read the plurality of data records in the distributed database;

means for receiving the plurality of data records read from the distributed database;

means for sending the request to perform the transaction to write the single data record dependent upon the plurality of data records read from the distributed database; and

means for receiving an acknowledgment in response to receiving the request to perform the transaction to write the single data record dependent upon the plurality of data records read from the distributed database.