Skip to content

Advanced Usage

Deeper operational recipes for clusters managed by this provider. Everyday deployment and management workflows live in Usage Examples.

Cluster name vs deployment name

The filesystem resources let you decouple the qfsd cluster name (shown in the Qumulo UI) from the prefix used to name cloud infrastructure. This applies to qumulo_filesystem_aws, qumulo_filesystem_azure, and qumulo_filesystem_gcp alike; the AWS resource is shown. Pick one of two shapes:

resource "qumulo_filesystem_aws" "example" {
  name = "prod-data" # qfsd cluster name AND infrastructure name seed
  # ... other required fields ...
  deletion_protection = true # recommended: guard the cluster's EC2 instances and S3 buckets

  timeouts {
    create = "90m"
    delete = "30m"
  }
}

name is limited to 2-15 characters (the qfsd cluster-name limit). It sets the cluster name as-is and seeds cloud resource names in lowercase. This is the default, back-compatible shape.

resource "qumulo_filesystem_aws" "example" {
  cluster_name    = "prod-data"     # qfsd cluster name (case preserved)
  deployment_name = "acmestorage01" # seeds cloud resource names (lowercase)
  # ... other required fields ...
  deletion_protection = true # recommended: guard the cluster's EC2 instances and S3 buckets

  timeouts {
    create = "90m"
    delete = "30m"
  }
}

cluster_name keeps the 2-15 character qfsd limit. deployment_name may be longer, with a per-cloud cap:

  • AWS: up to 36 characters
  • Azure: up to 15 characters (capped by Azure's tight storage-account name limit)
  • GCP: up to 16 characters (capped by GCP's tight service-account name limit)

The name shortcut also sets the cluster name, so it cannot exceed 15 characters; a deployment_name longer than 15 requires this pair.

Set name alone, or both cluster_name and deployment_name; mixing name with either is rejected.

The real resource names carry a random suffix

On first deploy the provider appends an immutable random suffix to deployment_name to form deployment_unique_name (for example acmestorage01-h12g9v). That unique name, not the bare deployment_name, is what actually names and tags every cloud resource and every cluster node. It stays fixed for the deployment's life. Automation that needs the real resource names should read the computed deployment_unique_name attribute rather than deployment_name.

Inspecting EC2 startup scripts

AWS stores this provider's EC2 user data as gzip-compressed, base64-encoded content. To inspect the rendered startup script for a cluster node or provisioner instance, replace the instance ID and run:

aws ec2 describe-instance-attribute \
  --instance-id i-0123456789abcdef0 \
  --attribute userData \
  --query 'UserData.Value' \
  --output text \
| base64 --decode \
| gunzip