IBM E02HMLL-I Implementation Guide - Page 160

If you have a need to retrieve information from an application, but do not want to use an adapter because of the impact on performance of sending a business object request to the application and receiving the response, you have several alternatives: v You can design a custom business object that only has attributes for the fields you need to affect in the operation. For instance, you may have a large application-specific business object designed for an application entity involved in a particular business process. If you need to retrieve just a small subset of information about that entity as part of another business process, you might be concerned about the performance impact of retrieving the full business object for just those few fields. To minimize the impact, you can make a copy of the application-specific business object and reduce its structure to only those fields that are required. That way only a small amount of information must be processed. This approach is best for retrieving information in the application database that changes frequently, because each request will retrieve information that is current at the time of the request. v You can replicate the application information on the same server that hosts the InterChange Server Express database. You can use the utilities provided by the database vendor to script the structure and data of a table into a file and then reproduce the table and its data in a database other than the one that the application itself uses. This approach is most useful for small amounts of information that is flat and static, such as a lookup table. You should not, however, use this approach for large entities that span multiple tables because the queries involved would be difficult to create and maintain, whereas using business objects to represent the entity would be easy for you to develop and for others to maintain. It is also not very good for tables that are very dynamic and have new records added frequently, because you either must manually update the replicated table frequently, or risk the interface having outdated information. Creating database connection pools and database connections A database connection pool consists of a number of reserved database connections. The reserved database connections are made available only to the collaboration and map processes that you design to use the pool. To create a database connection pool, you define the values necessary for making a database connection beforehand. The integration system saves this database connection information and uses it at runtime to more quickly establish connections for collaboration and mapping processes that you have assigned. The configuration of database connection values that you define can be used by one or more pools. For each pool you will specify a number of connections; these connections will be allocated, used, and freed back to the pool. Note: It is recommended that you connect to the InterChange Server Express instance that will use the database connection pool in System Manager before you create the pool, so that you can validate the connection. Do the following to create a database pool: 1. Right-click the Database Connection Pools folder in your integration component library in System Manager and choose Create New Database Connection from the context menu. 148 IBM WebSphere Business Integration Server Express and Express Plus: System Implementation Guide