You can leverage the Kafka partitions with Siebel CRM Event Publication and Subscription to improve scalability, support higher volumes and enhance the speed of data exchange through parallelism. Partitions increase scalability of data exchange by distributing messages within a topic to multiple Kafka servers.

To utilize the benefits provided by Kafka partitions, topics can be partitioned, and then messages within a topic can be spread over a number of "buckets" across multiple Kafka servers. These “buckets” (partitions) together comprise a topic. The distribution of data to different partitions allows multiple consumer applications to read data from multiple partitions at the same time, thereby consuming data of the topic faster. Similarly, applications using multiple producer threads can write data to multiple Kafka partitions for faster data delivery. Data can be sent targeted to specific partitions within a topic, thereby allowing different types of events to be written to the same Kafka topic at the same time while keeping a logical separation. Kafka guarantees that any consumer of a given topic-partition will always read that partition's events in exactly the same order as they were written.

You must configure different parameters in the configuration files to leverage Kafka partitions, these are discussed in the appropriate configuration files sections later in this document.

Note: Using partitions with Siebel CRM Event Publication and Subscription is optional, but if you chose to use partitions, you must plan partitioning of data at design-time for more streamlined and efficient integration strategies.