Data integrity in message flows

Subscription data that is retrieved from client applications (for example, topics from publishers and subscribers, and content filters from subscribers), and the character data entered through the IBM® Integration Toolkit (for example, message flow names), are stored by the integration node. This data is translated from its originating code page to the code page of the process in which the integration node is running, then by the database manager to the code page in which the database or databases were created.

To preserve data consistency and integrity, ensure that all this subscription data and IBM Integration Toolkit character data is originated in a compatible code page to the two code pages to which it is translated. If you do not do so, you might get unpredictable results and lose data.

The restrictions described do not apply to user data in messages. Ensure that any data in messages generated by your applications is compatible with the code page of all databases that you access from your message flows.

SQL statements that are generated as a result of explicit reference to databases in message processing nodes can contain character data that has a variety of sources. For example, the data might have been entered through the IBM Integration Toolkit, derived from message content, or read from another database. All this data is translated from its originating code page to the code page in which the integration node was created, then by the database manager to the code page in which the database was created. Ensure that these three code pages are compatible to avoid data conversion problems.