Skip to main content

Start, Stop or Destroy Clusters

You can easily start, stop or destroy existing clusters, both from the dashboard and from the CLI. Each action has certain consequences for your jobs and data on the cluster:

  • Stop: All jobs are cancelled and the cluster is stopped. Any data on the cluster is persisted. You can not log into the cluster with SSH while it's stopped. All billing for the resource will be paused.
  • Start: The resource is re-provisioned. It will take anywhere from 1-5 minutes depending on the provider, the resource type, attached storage and more. Billing will be resumed, and buckets remounted (if any). You can track the status of the cluster from the dashboard, or with wisp cluster list.
  • Destroy: The resource, and all related resources are destroyed. Any data on the instance will be deleted. All billing stops. To run a new job, you need to run wisp run and provision a new cluster.

Example: Modifying clusters from the dashboard​

Go the the Dashboard and you should be able to see existing clusters under "Current Instances". To modify it's state, click the dropdown menu (three dots) under Actions and select Start / Stop or Delete.

After clicking the button, Wisp will start modifying the state of the cluster. This may take some time.

Example: Modifying clusters from the CLI​

First identify what cluster you need to modify by one of the following methods:

  • Run wisp cluster list and identify the relevant cluster
  • Go to the dashboard and copy the cluster name from the cluster overview.

Then execute one of the following commands:

# Stop the cluster
wisp cluster stop <CLUSTER_NAME>
# Start the cluster
wisp cluster start <CLUSTER_NAME>
# Destroy the cluster
wisp cluster destroy <CLUSTER_NAME>

All of the commands takes an additional --wait flag, the keeps the command active until the modification has completed.