Getting started with IBM Business Automation Insights

Use the following procedure to prepare your IBM Cloud Private cluster for a Business Automation Insights deployment.

1. Install IBM Event Streams in your IBM Cloud Private cluster. You can use the IBM Event Stream Community Edition from IBM Cloud Private catalog.

   IBM Event Streams Foundation V2018.3.0 for Linux on x86 64-bit is also available in the download from IBM Passport Advantage®. The part number is CNW2DML.

2. Define the Kafka topics that are needed for Business Automation Insights.

   Create the necessary Kafka topics, see Using IBM Event Streams as your Kafka provider.

    New in 18.0.2  If you do not create the Kafka topics, IBM Business Automation Configuration Container does it for you. However, the automatically created topics probably do not meet the needs of a production environment.

   Role	Description	Required
   Ingress	Processing jobs consume raw events that are sent by emitters.	Yes
   Egress	Processing jobs produce time series and summarized entities.	Optional
   Service	Required by administration service to broadcast service messages to processing jobs.	Yes

   Tip: Specify the role in the topic names to avoid confusion. If you create the Kafka topics manually, set the appropriate degree of parallelism and number of replicas to match your production requirements.
3. Configure storage for Apache Flink. Refer to Setting up Apache Flink Storage.
4. Prepare to use the embedded Elasticsearch and Kibana environment.
   1. Create a persistent volume (PV) for every pod that stores Elasticsearch data. Refer to Setting up Elasticsearch storage.
   2. Ensure that the target namespace is bound to a pod security policy that supports running privileged containers. Refer to Modify your IBM Cloud™ Private security policy.
5. The use of Hadoop Distributed File System (HDFS) is optional. It is not used in this Getting started. HDFS can be enabled as a post-installation step. For more information, see Advanced updates.

Use the following procedure to install Business Automation Configuration Container on your IBM Cloud Private cluster.

1. Prepare to install IBM Business Automation Configuration Container.
2. Load your purchased container images.
3. Install the configuration container.

Use the following procedure to install Business Automation Insights on your IBM Cloud Private cluster.

1. In the Business Automation Configuration Container interface, enter your deployment name and description.

   Launch the Business Automation Configuration Container and enter the name of the release that you want to give it, for example Getting started with BAI.

2. Select the platform where you want to deploy Business Automation Insights.

   Set the platform configuration parameters for your Business Automation Insights platform: your master node in IBM Cloud Private, the namespace of the Business Automation Configuration Container release, and its persistent volume claim (PVC).

3. Select IBM Business Automation Insights as the product that you want to configure and deploy.
4. Choose and configure the product components that you want to deploy.

   Common Settings

   Helm release name sets the name of the release.

   Kakfa Cluster Settings sets the Kafka topic names.

   Apache Kafka set

   Set Bootstrap servers, Username, Password and Server certificate to configure IBM Business Automation Insights with IBM Event Streams as your Kafka provider. Take note of the Kafka bootstrap server URLs, download the Kafka server certificate and truststore for the SSL connection, and generate an API key. Refer to Using IBM Event Streams as your Kafka provider.

   For more information, see https://ibm.github.io/event-streams/getting-started/client/#securing-the-connection.

   Install Elasticsearch and Kibana

   Choose port numbers to expose the Elasticsearch REST endpoint (31200) and Kibana UI (31501), or keep fields empty for ports to be dynamically allocated. Go to Elasticsearch Settings > Client Node Settings > External port, and Kibana Settings > External port.

   Set storage values according to the persistence volumes (PV) you previously prepared. Go to Elasticsearch Settings > Data Node Settings > Storage Settings.

   Administration Service

   Choose a port number (31100) for the External port to expose the Administration service REST endpoint, or keep the field empty for the port to be dynamically allocated.

   Event Processing Jobs

   Review and select the event processing jobs that you want to deploy. For more information, see IBM Business Automation Insights overview. By default, all of the event processing jobs are installed. You can choose to disable processing jobs you do not need. It is possible to install processing jobs post installation by Updating a Business Automation Insights deployment.

   Apache Flink

   Set the Persistent Volume Claim values according to the PV you previously prepared.

5. Review your configuration in the Business Automation Configuration Container user interface before you go to the final Deploy step.
6. Click Deploy and follow the progress. For more information, see Reviewing and deploying.

Use the following procedure to verify the Business Automation Insights deployment on your IBM Cloud Private cluster.

1. From the IBM Cloud Private administration console, click Menu > Workloads > Helm releases.
2. Search for the name that you entered for your Business Automation Insights release, for example bai.
3. Check whether your release name is marked Deployed.
4. In the Pod table, check that for each pod the status is shown as Completed or Running, with all the containers ready. For example, the READY column shows 2/2.
5. In the Job table, check that the jobs are successful.
6. To open the Business Automation Insights services, use the Launch menu.
   You can also find the URIs and credentials in the Notes section of the page.

   Note: For the administration REST API service, append /api/health to the URL to check its status.