About database storage areas

Database storage areas (like file storage areas) provide permanent storage for document content.

Managed by Content Engine, a database storage area converts document content into Binary Large Objects (BLObs) for storage in the database specified as the object store database.

When you create an object store, the wizard prompts you to specify the default content storage: either a database storage area or a file storage area. Choosing the file storage area does not eliminate the database storage area; this selection simply determines which content store serves as your default store when you add documents to Content Engine. Enterprise Manager automatically creates a database storage area (contained by a table named "Content" ) when you create an object store and specify the default storage area for document content is a database.

Note that each object store has only one database storage area.

When choosing either file storage or database storage for a document class, consider all the factors. Because of Java™ Database Connectivity (JDBC) driver limitations, consider storing large content in file storage.

Determine the maximum size of the content elements your users store. This affects setting up database storage areas or file storage areas. When you create an object store, a database storage area is provided by default, allowing you to store content as database BLObs. You can also create one or more file storage areas to store content on local or remote file systems. If your users store large individual documents or other content elements, and you create the associated object stores on Microsoft SQL Server or IBM DB2 databases, use only file storage areas. Otherwise, users can encounter memory-related errors when retrieving or indexing the large content.

NOTE  Controlled tests with limited concurrency exhibited errors when run with files that were 300 MB or larger. Factors affecting this file-size limitation include driver and application server memory demands, other activity such as concurrent retrieval or indexing of large content, and JVM memory allocations.

NOTE   This is not an issue with Oracle where, during commit actions, content is streamed, rather than stored, in memory.

See Document content and Concepts: object stores for more information.