Storage Volume Definitions
Persistent volumes must be created before they are referenced in a workflow. You can create them using the CLI, web UI, or API. Below is the CLI method — you specify the provisioner name and the volume name as arguments to the create command.
$ fuzzball volume create shared-nfs my-dataset
Volume "my-dataset" created on provisioner "shared-nfs"The first argument is the
provisioner name and the second is the
volume name. The provisioner must exist and your group must have create access on it.
Optional flags allow setting additional properties at creation time:
| Flag | Description |
|---|---|
--description | Human-readable description |
--size | Requested volume capacity (e.g., 10GB) |
--owner | Override volume owner UID |
--permissions | Override directory permissions |
# fuzzball volume create shared-nfs project-data --description "Shared research data" --size 50GB
Volume "project-data" created on provisioner "shared-nfs"To create a volume, you must be logged into a user context and have a group selected. Your group must be listed in the provisioner’screatepolicy. See Access Policies for details.
Ephemeral volumes are created automatically when a workflow runs and destroyed when the workflow completes. They do not need to be created in advance. See Storage Volumes and Workflows for details on how to define ephemeral and persistent volumes in workflows.
Persistent volumes must always be created before they are referenced in a workflow. Workflows that reference a persistent volume that does not exist are rejected at submission time.