Administration commands restore, cleanup and set-password in dremio-admin needs to be run when the Dremio cluster is not running. So, before running these commands, you need to shutdown the Dremio cluster.
As part of the Helm chart, we support invoking the dremio-admin commands via a dremio-admin pod. Consult the Admin CLI documentation for Dremio for a complete list of dremio-admin commands that can be invoked.
Starting Dremio Admin Pod
The dremio-admin pod is created via the Helm chart. During this process, Dremio will become unavailable to end users as the other pods are shutdown during this process.
To invoke the dremio-admin pod, use the following Helm command:
$ helm upgrade <release-name> dremio_v2 --reuse-values --set DremioAdmin=trueStopping Dremio Admin Pod
To stop the dremio-admin pod and restart the other Dremio pods, use the following Helm command:
$ helm upgrade <release-name> dremio_v2 --resuse-values --set DremioAdmin=falseConnecting to the Dremio Admin Pod
Once you have started the dremio-admin pod, you can use the following command to access the pod:
$ kubectl exec -it dremio-admin -- bashThe above command will connect you to the dremio-admin pod. Once there, you can invoke the dremio-admin commands normally from within the pod.
Copying Files
To copy contents from the dremio-admin pod, you can use the following command:
$ kubectl cp dremio-admin:<directory> <local-directory>For example, to copy the contents of the Dremio db directory to a db_backup directory on your local machine, you can do the following:
$ kubectl cp dremio-admin:data/db db_backup