Workspaces

A workspace is an isolated environment for a team or project - its own catalogs, clusters, access control, and run history.

A workspace is an isolated environment within e6data. Each workspace has its own:

• Catalogs (metastores you've connected)
• Clusters (compute that runs queries)
• Access control (users, roles, and service accounts)
• Run history

Workspaces are the unit of isolation in e6data. You typically have one workspace per environment (for example, prod, staging) or per team.

Workspace types

When you create a workspace, you pick a deployment model:

Type	What it means
Serverless	e6data manages the compute environment. Your data stays in your cloud.
In Your VPC	Compute runs inside your own Kubernetes cluster.

See Deployment models for the differences.

In this section

• Create and manage workspaces - set up and update a workspace
• Workspace settings - enable or disable a workspace
• Delete workspace - permanently remove a workspace

Choosing how many workspaces to create

Common patterns:

Pattern	Workspaces
Environment isolation	One per environment: prod, staging, dev
Team isolation	One per team or product line
Tenant isolation (SaaS)	One per customer tenant

There's no hard limit on workspace count. Each is independently billed for compute, and access control is per-workspace.

Workspace lifecycle

[Provisioning] → [Running] → [Deleted]
                     ↓
                 [Disabled]
                     ↓
                 [Running]

• Provisioning - initial setup
• Running - normal operation
• Disabled - administrative pause; resume with Enable
• Deleted - permanently removed; not recoverable

Disable vs delete

Action	Purpose	Reversible	Compute cost
Disable	Administrative pause	Yes - enable	None
Delete	Permanent removal	No	None (resource gone)

Disable for compliance holds or offboarding when you might bring the workspace back; delete only when you're certain you won't. To pause compute for cost savings, suspend individual clusters instead.