Event Processing is an operator-based release and uses custom resources to define the deployment configuration. Event Processing requires the installation of the IBM Operator for Apache Flink and the Event Processing operator.
These operators deploy and manage the entire lifecycle of your Flink and Event Processing instances.
- Flink instances are defined by the
FlinkDeployment
custom resource type. - Event Processing instances are defined by the
EventProcessing
custom resource type.
A deployment of Event Processing consists of the IBM Operator for Apache Flink and the Event Processing operator, along with an instance of Flink and an instance of Event Processing.
Note: Ensure that the Event Processing version and the IBM Operator for Apache Flink version are the same.
Installing Event Processing includes the following high-level steps:
-
Depending on your platform and environment, select the right procedure:
- Red Hat OpenShift Container Platform online environment which has access to the internet.
- Offline environment (also referred to as air-gapped or disconnected).
- Other Kubernetes platform online environment which has access to the internet.
- Install the IBM Operator for Apache Flink operator. The operator will then be available to install and manage your Flink instances.
- Install the Event Processing operator. The operator will then be available to install and manage your Event Processing instances.
- Using the IBM Operator for Apache Flink operator, install an instance of the Flink by applying the
FlinkDeployment
custom resource type. - Using the Event Processing operator, install an instance of the Event Processing by applying the
EventProcessing
custom resource type.