System and method for ordering a database flush sequence at transaction commit

• US 7,590,639 B1
• Filed: 04/29/2004
• Issued: 09/15/2009
• Est. Priority Date: 04/29/2004
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP AG (SAP SE)
• Inventors: Ivanova, Viktoriya Ts., Pavlov, Vladimir K.
• Primary Examiner(s): Fleurantin; Jean B
• Assistant Examiner(s): Myint; Dennis
• Application Number:
  US10/837,484
• Time in Patent Office: 1,965 Days
• Field of Search: 707/203, 707/100, 707/2
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2365 Ensuring data consistency and integrity
• CPC Class Codes: G06F 16/2379 Updates performed during online database operations; commit processing
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing

Abstract

A system and method are described for ordering a database flush sequence prior to committing transaction changes to a database. In one embodiment, a plurality of data operations modifying persistent data objects are performed in a sequential order within an object-oriented environment. Relational dependencies between the persistent data objects are evaluated and the changes to the data objects are reordered based on the relational dependencies. Once reordered, the changes are committed to the database. In one embodiment, a depth first search algorithm is employed to determine the appropriate order to commit the data to the database.

Claims

1. A computer implemented method comprising:

performing a plurality of data operations of a transaction in a first sequential order, the transaction yet to be flushed to a relational database, the plurality, of data operations modifying a plurality of persistent data objects within an object-oriented environment,

evaluating a state property associated with the modified persistent data objects to determine to flush the transaction to the relational database;

evaluating relational dependencies between the modified persistent data objects within the object oriented programming environment prior to flushing the transaction to the relational database, the evaluating relational dependencies is based on a plurality of database integrity rules identifying conditions under which a series of operation are committed to the relational database and includes

evaluating the modified persistent data objects as an oriented graph, with vertices of the oriented graph representing tile modified persistent data objects and edges representing the relationships between the modified persistent data objects, and

identifying relational dependencies between the modified persistent data objects by performing a topological sort for traversing the oriented graph and for any two persistent data objects, A and B, data object A depends on persistent data object entity B if persistent data object A has a foreign key to persistent data object B, and B is created in the transaction or if persistent data object has a foreign key to persistent data object A, and A is deleted in the transaction;

reordering the plurality of data operations of the transaction within the object-oriented programming environment after evaluating the relational dependencies between the modified persistent data objects to a second sequential order based on the relational dependencies; and

flushing the transaction to the relational database, the flushing includes committing the data operations of the transaction to the relational database in the second sequential order in accordance with the reordering.

2. The method as in claim 1 wherein the modified persistent data objects are entity beans managed by an EJB container.

3. The method as in claim 2 wherein each entity bean is associated with a particular database table and wherein each of the data operations comprise either an insert command, inserting a new row into a database table or a delete command, removing a row from a database table.

4. The method as in claim 3 wherein relational dependencies are described in entity bean relationship fields, the relationship fields representing mappings between primary keys and foreign keys within the relational database.

5. The method as in claim 1 wherein evaluating relational dependencies comprises identifying modified persistent data objects which depend on other modified persistent data objects and wherein reordering the plurality of data operations comprises ensuring that any modified persistent data objects which depend on other modified persistent data objects are flushed only after the modified persistent data objects on which they depend are flushed.

6. The method as in claim 1 further comprising:

implementing the topological sort by executing a depth first search algorithm on the oriented graph.

7. A system comprising:

A relational database implemented in a data storage device, the relational database coupled to an object-oriented programming environment;

A persistence management layer within the object-oriented programming environment to

perform a plurality of data operations of a transaction in a first sequential order, the transaction yet to be flushed to a relational database, the plurality, of data operations modifying a plurality of persistent data objects,

evaluate a state property associated with the modified persistent data objects to determine to flush the transaction to the relational database;

evaluate relational dependencies between the modified persistent data objects within the object oriented programming environment prior to flushing the transaction to the relational database, wherein evaluating of relational dependencies is based on a plurality of database integrity rules identifying conditions under which a series of operation are committed to the relational database and includes

evaluating the modified persistent data objects as an oriented graph, with vertices of the oriented graph representing tile modified persistent data objects and edges representing the relationships between the modified persistent data objects, and

identifying relational dependencies between the modified persistent data objects by performing a topological sort for traversing the oriented graph and for any two persistent data objects, A and B, data object A depends on persistent data object entity B if persistent data object A has a foreign key to persistent data object B, and B is created in the transaction or if persistent data object has a foreign key to persistent data object A, and A is deleted in the transaction;

reorder the plurality of data operations of the transaction within the object-oriented programming environment after evaluating the relational dependencies between the modified persistent data objects to a second sequential order based on the relational dependencies; and

flush the transaction to the relational database, the flushing includes committing the data operations of the transaction to the relational database in the second sequential order in accordance with the reordering.

8. The system as in claim 7 wherein the modified persistent data objects are entity beans managed by an EJB container.

9. The system as in claim 8 wherein each entity bean is associated with a particular database table within the relational database and wherein each of the data operations comprise either an insert command, inserting a new row into a database table or a delete command, removing a row from a database table.

10. The system as in claim 9 wherein relational dependencies are described in entity bean relationship fields, the relationship fields representing mappings between primary keys and foreign keys within the relational database.

11. The system as in claim 7 wherein evaluating relational dependencies comprises identifying modified persistent data objects which depend on other modified persistent data objects and wherein reordering the plurality of data operations comprises ensuring that any modified persistent data objects which depend on other modified persistent data objects are flushed only after the modified persistent data objects on which they depend are flushed.

12. The system as in claim 7 wherein the transaction reordering logic implements the topological sort by executing a depth first search algorithm on the oriented graph.

13. An article of manufacture comprising a machine-readable storage medium, including program code which, when executed by a machine, causes the machine to perform the operations of:

performing a plurality of data operations of a transaction in a first sequential order, the transaction yet to be flushed to a relational database, the plurality, of data operations modifying a plurality of persistent data objects within an object-oriented environment,

evaluating a state property associated with the modified persistent data objects to determine to flush the transaction to the relational database;

evaluating relational dependencies between the modified persistent data objects within the object oriented programming environment prior to flushing the transaction to the relational database, the evaluating relational dependencies is based on a plurality of database integrity rules identifying conditions under which a series of operation are committed to the relational database and includes

evaluating the modified persistent data objects as an oriented graph, with vertices of the oriented graph representing tile modified persistent data objects and edges representing the relationships between the modified persistent data objects, and

identifying relational dependencies between the modified persistent data objects by performing a topological sort for traversing the oriented graph and for any two persistent data objects, A and B, data object A depends on persistent data object entity B if persistent data object A has a foreign key to persistent data object B, and B is created in the transaction or if persistent data object has a foreign key to persistent data object A, and A is deleted in the transaction;

reordering the plurality of data operations of the transaction within the object-oriented programming environment after evaluating the relational dependencies between the modified persistent data objects to a second sequential order based on the relational dependencies; and

flushing the transaction to the relational database, the flushing includes committing the data operations of the transaction to the relational database in the second sequential order in accordance with the reordering.

14. The article of manufacture as in claim 13 wherein the modified persistent data objects are entity beans managed by an EJB container.

15. The article of manufacture as in claim 14 wherein each entity bean is associated with a particular database table and wherein each of the data operations comprise either an insert command, inserting a new row into a database table or a delete command, removing a row from a database table.

16. The article of manufacture as in claim 15 wherein relational dependencies are described in entity bean relationship fields, the relationship fields representing mappings between primary keys and foreign keys within the relational database.

17. The article of manufacture as in claim 13 wherein evaluating relational dependencies comprises identifying modified persistent data objects which depend on other modified persistent data objects and wherein reordering the plurality of data operations comprises ensuring that any modified persistent data objects which depend on other modified persistent data objects are flushed only after the persistent data objects on which they depend are flushed.

18. The article of manufacture as in claim 13 comprising additional program code to cause the machine to perform the operations of:

implementing the topological sort by executing a depth first search algorithm on the oriented graph.