Customizing persistent storage

Edit online

IBM Business Automation Insights is based on the IBM Cloud Pak® for Business Automation environment, which by default deploys persistent volumes and persistent volume claims (PVCs). You can opt for static provisioning by creating persistent volumes manually.

About this task

You cannot remove the default persistent volumes (PV) and persistent volume claims (PVC) but you can customize them for your business needs.

By default, the volumes are named after your custom resource name. For information about the values required for high availability, see Replica configuration for high availability and fault tolerance.

The following table summarizes persistent storage for IBM Business Automation Insights.

Table 1. Persistent storage
Volume Default volume name Default storage Access mode Number of volumes
Flink volume <CR_NAME>-bai-pvc 20Gi ReadWriteMany 1
OpenSearch master-data data-opensearch-ib-6fb9-es-server-all-<replica>
  • 20Gi for a demo (evaluation) configuration
  • 50Gi for an enterprise configuration
 ReadWriteOnce 3
OpenSearch snapshot opensearch-ib-6fb9-es-server-snap
  • 5Gi for a demo (evaluation) configuration
  • 10Gi for an enterprise configuration
 ReadWriteMany Only one, regardless of the number of replicas
The Flink volume is used by multiple pods for normal operation. You configure persistent volume claims (PVCs) by setting parameter values, as documented in Persistent Volume Claim.

Procedure

  1. Log in to your Kubernetes cluster.
    On OpenShift, use the oc tool.
    oc login https://<CLUSTER_IP>:8443
  2. Go to the namespace where you plan to deploy Business Automation Insights. 
    oc project <NAMESPACE>
  3. If you need static provisioning rather than the default dynamic provisioning, create the persistent volumes manually.
    Business Automation Insights requires some persistent volumes. For example, Apache Flink needs a persistent volume to store its internal state and to support fault tolerance and high availability.