Managing clusters

You can manage your Coiled clusters in two ways: using the coiled Python package or via Coiled’s web interface.

The video below will walk you through creating custom Coiled clusters and managing your running clusters.

Listing and deleting clusters

The coiled.list_clusters() method will list all active clusters:

coiled.list_clusters()

Note that when a cluster is created, by default, a unique name for the cluster is automatically generated. You can provide your own cluster name using the name= keyword argument for coiled.Cluster.

coiled.delete_cluster() can be used to delete individual clusters. For example:

coiled.delete_cluster(name="my-cluster")

deletes the cluster named “my-cluster”.

Note

Listing and deleting only work for active clusters. Your account dashboard will show all the clusters that have been created, but their status will show as stopped.

Web interface

Coiled maintains a web interface where you can, among other things, view your recently created Dask cluster along with other information like how many workers are in the cluster, how much has running the cluster cost, etc. For more information, see https://cloud.coiled.io/clusters.

../_images/clusters-table.png

The video below walks you through the Coiled dashboard: