Explanation
Why the waste happens and who it affects.
When no auto-termination policy is configured, these clusters remain active indefinitely, incurring unnecessary charges for both Databricks and cloud infrastructure usage. This inefficiency is especially common in interactive clusters that are user-managed, ephemeral, or exploratory in nature. While Databricks provides built-in support for cluster auto-termination, teams often overlook it unless it's enforced through workspace policies. Without this safeguard in place, idle clusters can persist unnoticed for hours or days, leading to avoidable cost.
Billing model
The pricing dimensions that drive this cost.
Databricks clusters accrue cost per second through:
- Databricks Unit (DBU) charges
- Vary by workload type (interactive, job, SQL)
Underlying cloud compute - billed through the host cloud provider (e.g., EC2, Azure VMs) Clusters without auto-termination continue to run - and generate cost - even if idle or abandoned.
How to detect
4 checks to find it in your estate.
- Identify clusters that do not have auto-termination enabled
- Check for clusters with long idle times and no active workloads
- Analyze cost reports to detect charges from underutilized or inactive clusters
- Review workspace-level cluster policies and defaults to ensure consistent enforcement
How to fix
4 ways to remove the waste.
- Enable auto-termination for all clusters that do not require persistent runtime
- Set cluster policies to require auto-termination configuration for new clusters
- Establish reasonable inactivity thresholds based on workload type (e.g., 30-60 minutes for interactive)
- Educate users on the financial impact of idle clusters and the role of auto-termination as a cost control mechanism
Documentation
Vendor references for pricing and configuration.