Google Cloud Supported Services
NightOps supports scheduling for Google Cloud compute services.
Supported Services
Compute Engine VMs
Stop and start Compute Engine virtual machines on schedule. All persistent disks and configuration are preserved. External IPs may change unless you use static IPs.
- Standard and custom machine types
- Preemptible VMs are supported but may be terminated by Google
- Instances in managed instance groups (handled separately)
Cloud SQL Instances
Stop and start Cloud SQL database instances. All data is preserved. Unlike AWS RDS, Google Cloud SQL does not have a maximum stop period.
- MySQL, PostgreSQL, and SQL Server
- Single-zone and high-availability configurations
- Read replicas stop automatically with the primary
Cloud Run Services
Scale Cloud Run services to zero instances during off-hours. Since Cloud Run scales to zero by default, this is primarily useful for services configured with minimum instances.
GKE Node Pools
Scale GKE node pools to zero nodes during off-hours. Your cluster control plane remains running. This works with both Standard and Autopilot clusters (for Standard mode node pools).
Managed Instance Groups
Set managed instance group target size to zero during off-hours and restore to your configured size when needed.
Quick Start
- Label your resources: Add
nightops-managed: trueto any resources you want NightOps to manage. - Set up a Service Account: Create a service account with the required IAM roles.
- Connect your project: Add your GCP project in the NightOps dashboard using the service account key or Workload Identity.
- Create a schedule: Define when your resources should be running.
Labeling Best Practices
We recommend using consistent labels across your resources:
| Label | Value | Purpose |
|---|---|---|
nightops-managed | true | Required for NightOps to manage the resource |
nightops-environment | dev, staging, etc. | Group resources by environment |
nightops-production | true | Protect resource from any scheduled operations |
Pricing Impact
When NightOps stops your GCP resources, you stop paying for compute. You continue to pay for:
- Persistent disks attached to stopped Compute Engine VMs
- Cloud SQL storage while instances are stopped
- GKE cluster management fee (fixed cost regardless of nodes)
- Static external IP addresses not attached to running resources
Next Step
Set up the required IAM permissions: Google Cloud IAM Policy →