Different types of data stores are available in ibaPDA for different purposes and methods of data storage. Depending on the licenses registered in the license container, different types of data stores are available for configuration in the dialog.

This documentation describes the "Kafka cluster timebased data store" type of recording. This type of data store writes timebased data into a Kafka cluster.

The architecture of Apache Kafka consists of a cluster computer network. So-called ‘brokers’ in this computer network store received messages with a time stamp in so-called ‘topics.’

Topics can be split into several partitions in the cluster and replicated. Applications that write messages in a Kafka cluster are called producers. The messages are sent to the leader. Among themselves the brokers take care of the replication with the followers. Consumers are applications that read messages from the Kafka cluster.

ibaPDA has the role of a producer in a Kafka environment: The data store sends messages to the cluster.

Additional information on Apache Kafka can be found at

• https://kafka.apache.org/

• https://www.confluent.io/what-is-apache-kafka

Chapter Signal selection describes the selection of the signals that are to be recorded in the topics.

The data can be continuously recorded or recorded by trigger, see chapter Trigger mode.