Every action you can take in the DAAKYI console — creating a VM, attaching a volume, taking a snapshot, updating a firewall rule — is also a REST API call. That parity is the foundation for automating your infrastructure without any special tooling.
Pattern 1: Provision from Your Pipeline
Call the VM and network APIs directly from CI/CD jobs (GitHub Actions, GitLab CI, Jenkins) using a scoped API key. A pipeline stage can provision a fresh VM, deploy, run smoke tests, and tear the VM down — paying compute only for the minutes it ran.
Pattern 2: Snapshot Before Every Release
Add a pre-deploy step that snapshots your application and database VMs through the snapshot API. If a release goes wrong, roll back to a known-good state in minutes instead of rebuilding.
Pattern 3: Scheduled Operations on Worker VMs
Run cron-driven scripts on a small worker VM for recurring jobs — nightly report generation, log archival to DC Object Storage, powering off dev environments outside working hours. Component-based billing means powered-off VMs stop billing for compute.
Best Practices
- Use scoped API keys per pipeline, never personal credentials
- Make scripts idempotent — check whether a resource exists before creating it
- Tag automated resources so metering reports show what your pipelines cost
- Wire DC Monitoring alarms to webhooks so failures page your incident tooling
- Review DC Audit Trails to keep every automated change attributable
