Configuring IBM Business Automation Insights

To install IBM® Business Automation Insights on IBM Cloud Private, you use Business Automation Configuration Container.

1. Provide the Common settings values.
   • Helm release name: Give a name to the release that will be deployed. You will need that name later in Helm chart commands, for example to update your IBM Business Automation Insights deployment.
   • Kafka Cluster Settings: Specify the Kafka topics that you created in Preparing your Kafka cluster.
   • Kerberos configuration: If you intend to secure communications to Kafka or HDFS, or both, by using Kerberos authentication, see Enabling Kerberos.
2. Secure communications to Apache Kafka.

   To determine which security properties to set depending on the type of security you enforce in your Kafka cluster, see Securing communications and data.

3. Install Elasticsearch and Kibana.

   Embedded Elasticsearch and Kibana

   If you choose to install Elasticsearch and Kibana with IBM Business Automation Insights, make sure that the requirements described in Preparing for Elasticsearch storage are met.

   • Elasticsearch Settings > Client Node Settings > Client Service type: See Exposing IBM Business Automation Insights services.
   • Kibana Settings: You can customize the name and password of the user account used by Kibana to access the Elasticsearch REST API. For more information, see Managing user accounts and credentials.

   External Elasticsearch and Kibana

   If you prefer to use your own Elasticsearch environment, provide a URL that points to your Elasticsearch REST endpoint. If your Elasticsearch REST endpoint is secured, see Securing communications to Elasticsearch.

4. Set up the administration service.
   • Administration API service type: See Exposing IBM Business Automation Insights services.
   • Service user account: You can customize the user name and password of the user account that is used to access the administration service. For more information, see Managing user accounts and credentials.
5. Optional: Select event processing jobs.

   By default, all events processing jobs that IBM Business Automation Insights supports are installed. You can disable the processing jobs that you do not need. You can later install the processing jobs that you disabled at initial configuration by updating your IBM Business Automation Insights system. For more information, see Updating IBM Business Automation Insights.

6. Set up Apache Flink.
   • RocksDB Options Set: See Table 1 of Apache Flink parameters.
   • Job Checkpointing Interval: See Checkpoints and savepoints.
   • HDFS: Storage Bucket URL: You must specify a URL only if you install the processing job for raw events or if you want other processing jobs to store data in HDFS. For more information, see Preparing to use HDFS. Parameters Storage batch size, Inactive bucket check interval, and Inactive bucket threshold are used only if HDFS is enabled when you set Storage Bucket URL.
   • Zookeeper replicas: For high availability, each replica of Zookeeper runs in a separate node. By default, 3 replicas are installed. With this number, the system can continue to work even if one node is down, as long as the remaining nodes have enough resources. You can still run IBM Business Automation Insights in an environment with fewer nodes.
     Attention: However, if you reduce the number of Zookeeper replicas, be aware that you lose the benefit of high availability and fault tolerance. For high availability and fault tolerance to be effective, it is imperative to have at least 3 Zookeeper replicas running.
7. Review your configuration, and then move on to subsequent Business Automation Configuration Container steps until you are ready to complete deployment.
8. Click Deploy.
9. Verify your IBM Business Automation Insights configuration.

   See section IBM Business Automation Insights of Reviewing and deploying.