The key to managing VoltDB clusters in Kubernetes is to let the Helm charts do the work for you. You can use helm commands to perform all basic activities for running a database. This chapter explains how to use helm commands to:
Start the cluster for the first time
Stop and restart the cluster
Resize the cluster
Pause and resume
Start multiple clusters within one Kubernetes namespace
Subsequent chapters explain how to manage the database once it is running, how to modify the database and cluster configuration, and how to upgrade the VoltDB software itself.
As described in Chapter 2, Configuring the VoltDB Database Cluster you can customize every aspect of the database and the cluster using Helm properties and the configuration can be as simple or as complex as you choose. But once you have determined the configuration options you want to use, actually initializing and starting the database cluster is a single command, helm install. For example:
$ helm install mydb voltdb/voltdb \ --values myconfig.yaml \ --set global.voltdbVersion=13.2.0 \ --set-file cluster.config.licenseXMLFile=license.xml \ --set cluster.clusterSpec.replicas=5
When running VoltDB in Kubernetes, it is important to always use the latest versions of the Volt Operator and custom resource definition (CRD). The first time you start a database in Kubernetes, Helm automatically uses the latest operator and CRD. Helm continues to use the latest operator on subsequent installs. However, it does not automatically update the CRD; you must do this manually. Therefore, if you have used Helm to run VoltDB before, it is best to update the CRD before starting a new database, as described in Section 5.3.2, “Updating the Custom Resource Definition (CRD)”.
Documentation