Updating your Business Automation Insights deployment

You can update your IBM® Business Automation Insights system by passing new configuration property values on the Helm command line.

1. Retrieve Business Automation Insights Helm chart.

   docker run -v "$(pwd)":/mnt mycluster.icp:8500/<my-namespace>/bai-ibacc-job:<bai-version-number> cp ../../product-helm-charts/bai.tar.gz /mnt

   Replace <bai-version-number> with 18.0.0, 18.0.1, or 18.0.2, as appropriate.
2. If necessary, delete the job submitter jobs.

   You need to delete the job submitters jobs if you are planning to update parameters that affect the execution of processing jobs: Apache Flink settings (including RocksDB settings), Kafka configuration options, Elasticsearch general settings, or Kerberos authentication settings. These properties start with flink.*, bpmn.*, ingestion.*, icm.*, kafka.*, settings.*, kerberos.*, or elasticsearch.*.

   1. To retrieve the job names, run the get command as follows.

      kubectl get jobs --selector=release=<release-name> --namespace <my-namespace> | grep -v setup

   2. Run the delete on each job in the list as follows.

      kubectl delete job <job-name> --namespace <my-namespace>

3. If necessary, delete the bai-setup job by running the delete job command.
   You need to delete the bai-setup job if you update the elasticsearch.url property to change the Elasticsearch instance used by your Business Automation Insights system.

   kubectl delete job <my-release>-bai-setup --namespace <my-namespace>

4. Run the Helm upgrade command with the new property values.
   The following example increases the number of parallel instances for the processing of BPMN events from 1 to 2.

   helm upgrade <release-name> bai.tar.gz --reuse-values --set bpmn.parallelism=2 --namespace <my-namespace> --tls

   Tip: Increasing the parallelism of a job also increases Business Automation Insights memory and CPU usage. For more information, see Scaling an IBM Business Automation Insights system.