A reliable production release depends on a proven set of checks, not a silver bullet. This guide walks through the HanCloud team's pre-release checklist so you can avoid common failure modes.
Connect to code repository
Start with your repository. Once HanCloud has access, every push triggers a traceable pipeline. Test a non-default branch first to confirm permissions and webhooks are configured correctly.
Zero-configuration build
HanCloud detects your framework and creates the build configuration, so most projects need no custom CI. For an unusual project layout, use hancloud.json to override the defaults only when needed—convention comes first.
Preview environment
Every pull request gets an isolated preview with realistic data. Share the link in review so product and design can test the real experience before merge instead of guessing from screenshots.
Progressive rollout and rollback
Production releases use progressive delivery by default: shift a small share of traffic first, watch key metrics, then proceed once the release is healthy. If an alert fires, roll back to the last healthy version in one click, usually within seconds.
Add observability
Logs, metrics, and distributed traces work out of the box. Define SLOs for critical paths on day one so monitoring exposes every regression immediately—observability is a release guardrail, not an afterthought.
Establish a security baseline
Manage secrets instead of hard-coding them, grant least-privilege access, encrypt data in transit, and enable audit logs by default. Security is foundational, not a switch to flip later.
Summary
Together, these six checkpoints create an observable, reversible, and dependable release path. Make them routine and shipping becomes a normal part of the day instead of a source of anxiety.