Meta Catalog database

Each catalog exists as a separate OpenEdge RDBMS database. The database is built and accessed through SQL commands.

When you have a large number of files in a catalog, you might want to consider the following options to increase performance:

• Start the catalog database with a higher value for the Blocks in Database Buffer (-B) parameter. The larger buffer allows more records to be in memory which reduces disk access.
• Start the catalog database with a higher value for the Minimum Clients per Server (-Mi) parameter. SQL servers can support multiple connections and reduce memory usage when this parameter is increased.
• Consider using Asynchronous Page Writers if there are a lot of updates to the catalog database.

A catalog database stores no data that cannot be extracted from the source code by the content builders. You can manually delete the database files and rebuild the same catalog from its definition. The content builders can extract two types of data:

• Meta data that you include in the source code as 4GL annotations.
• Data that can be read from the code without compiling it, information that you can parse from simple string comparison operations.

Because the content builders do not compile the source code, they cannot make substitutions. Content builders cannot expand preprocessors. Content builders cannot read and analyze the contents of included files. Content builders cannot expand a LIKE phrase to show the attributes a field inherits from another field. You might need to consider this behavior when creating search queries. For example, if you had to check all fields using the LONGCHAR data type, you might want the search query to also return all field records that have the LIKE attribute in the catalog.

A catalog database is specific to the workspace in which it is created. In order to add a source file into a catalog, you must first import that file into the workspace associated with the catalog.

Caution: The catalog database's schema is subject to change without notice in future releases.  Any custom code created against the current schema might not function in future releases.