DigitalOcean Supported Services
NightOps supports automated start/stop management for the following DigitalOcean services.
Supported Services Overview
| Service | Stop Method | Start Method | Data Preserved | Config Preserved |
|---|---|---|---|---|
| Droplets | Power Off | Power On | Yes | Yes |
| Managed Databases | Scale Down | Scale Up | Yes | Yes |
| Kubernetes | Scale to 0 | Restore Size | Yes (with PVC) | Yes |
Droplets
DigitalOcean Droplets can be powered off and on without losing configuration or attached volume data.
API Operations
| Operation | API Endpoint | Description |
|---|---|---|
| List | GET /v2/droplets | List all Droplets |
| Stop | POST /v2/droplets/{id}/actions (power_off) | Power off a Droplet |
| Start | POST /v2/droplets/{id}/actions (power_on) | Power on a Droplet |
| Status | GET /v2/droplets/{id} | Check Droplet status |
What's Preserved
- Block storage volumes and data
- Networking configuration
- Floating IPs (if assigned)
- Tags and metadata
- Backups and snapshots
Caveats
Powered-off Droplets still incur storage costs for the reserved disk space. Only compute costs are saved.
The Droplet's public IP is preserved on power off/on. However, if you destroy and recreate, a new IP is assigned.
Managed Databases
Managed Database clusters cannot be stopped, but can be scaled down to the smallest plan during off-hours.
API Operations
| Operation | API Endpoint | Description |
|---|---|---|
| List | GET /v2/databases | List all database clusters |
| Scale Down | PUT /v2/databases/{id}/resize | Reduce to smallest size |
| Scale Up | PUT /v2/databases/{id}/resize | Restore original size |
| Status | GET /v2/databases/{id} | Check cluster status |
Supported Engines
- PostgreSQL
- MySQL
- Redis
- MongoDB
What's Preserved
- All data
- Connection strings
- Users and permissions
- Trusted sources (firewall)
Caveats
Resizing a database cluster takes 5-15 minutes and involves a brief connection interruption.
There's a minimum cluster size that cannot be reduced further. Cost savings depend on the difference between your production and minimum sizes.
Kubernetes (DOKS)
DigitalOcean Kubernetes node pools can be scaled to zero while preserving the control plane.
API Operations
| Operation | API Endpoint | Description |
|---|---|---|
| List | GET /v2/kubernetes/clusters/{id}/node_pools | List node pools |
| Stop | PUT /v2/kubernetes/clusters/{id}/node_pools/{pool_id} | Scale to 0 nodes |
| Start | PUT /v2/kubernetes/clusters/{id}/node_pools/{pool_id} | Restore node count |
| Status | GET /v2/kubernetes/clusters/{id} | Check cluster status |
What's Preserved
- Cluster configuration
- Node pool definitions
- Kubernetes workload definitions
- Persistent Volume Claims (with DO Volumes)
Caveats
The Kubernetes control plane is free on DigitalOcean, but the cluster still exists even with zero nodes.
Workloads requiring persistent storage need DigitalOcean Volumes attached via PersistentVolumeClaims.
Scale-up takes 2-4 minutes for new nodes to become ready.
Unsupported Services
| Service | Reason |
|---|---|
| App Platform | No pause functionality; billing continues for reserved resources |
| Spaces | Object storage; no stop mechanism needed (pay per usage) |
| Load Balancers | Delete changes IP; requires reconfiguration |
| Functions | Serverless; no stop needed (pay per invocation) |
Next Steps
- Set up API Token — Configure DigitalOcean access