Introduction

Messages are routed and transformed by message flows within IBM® Integration Bus. As a message passes through the message flow, it is manipulated by message processing nodes that are often implemented in ESQL (Extended Structured Query Language). The message is stored in memory in the form of a message tree, which is created from the input message bitstream and serialized into an output bitstream by a parser.

When you process small messages, for example messages that are a few kilobytes in size, only a small amount of memory is needed to fully parse the message.

However, when you process large messages, the amount of memory and processing time that is required to store the message tree can quickly increase. The size of a message tree can vary substantially, and it is directly proportional to the size of the messages that are being processed: The larger the messages, the larger the message tree. The inefficient use of parsers, using the wrong parser, or even parsing when you do not need to, can all increase the amount of memory that is used by the integration node.

When the message size grows to megabytes, it becomes even more important to keep memory usage to a minimum. If all your memory is being used by processing large messages, then there might not be enough to maintain your expected level of service. Also, there might not be enough resources for other processes, such as deployment, to run smoothly.

The best practice information in the following sections share common approaches to solving common problems based on real customer environments. They do not provide a "one size fits all" solution. They assume that you have a basic understanding of IBM Integration Bus. As technology evolves and improved functionality is added to the product, new recommendations and advice might be added to the information in these documents.