Backup, restore, and high availability

The Cloud Pak for Data System architecture mitigates the risk of unplanned downtime or data loss.

Backup and restore

Cloud Pak for Data System relies on Cloud Pak for Data backup and restore capabilities. The backup options depend on the version of Cloud Pak for Data you have installed:

If you are using Cloud Pak for Data version 4.0, refer to Backing up Cloud Pak for Data volumes to a persistent volume claim or object store in Cloud Pak for Data documentation.

If you have NPS® installed:

You can back up with nzbackup to NFS mounted storage.

High availability

The Cloud Pak for Data System architecture is based on high availability. The base system comprises 8 nodes:

3 control nodes - If a control node fails, the workflow is automatically shifted to another control node, OpenShift automatically moves the data.
3 worker nodes - If all three fail or are unavailable, a control node can become a worker node.
2 unset nodes - these can also be used if there are failures.

System expansion is available, you can buy and add extra nodes and HA switches as described in Expanding Cloud Pak for Data System.