Operations
Rollback
Section titled “Rollback”Roll back an environment to any previous release. Because releases are immutable and variables are snapshotted, a rollback restores exactly what ran before.
slipway rollback api --env production # previous releaseslipway rollback api --env production --to 1.3.2 # a specific oneA rollback is a deployment with triggered_by: rollback, so it goes through the environment’s protection rules like any other (use deployment.force with care, it is recorded and visible to policies). Blue/green and canary strategies roll back automatically on failure.
Deployment diff
Section titled “Deployment diff”ReleaseService.DiffReleases compares two releases: artifact versions and digests, variables, commits between them, and linked tickets. Use it in approvals (“what am I approving?”) and incident review (“what changed?”).
Freeze calendar
Section titled “Freeze calendar”EnvironmentService.GetFreezeCalendar shows all freeze periods and deployment windows across environments, with who created them and whether they allow override. Create a freeze with CreateFreeze or in the UI.
Environment board
Section titled “Environment board”A matrix of services by environments showing the deployed release, status, health signals from metric providers and drift state. Use it as the answer to “what is running where?”.
Drift detection
Section titled “Drift detection”For Kubernetes infrastructure, Slipway compares the desired state of the last deployment to the live state in the cluster and records a drift report (EnvironmentService.CheckDrift, ListDriftReports). Drift is surfaced on the board and can trigger notifications. Terraform and OpenTofu infrastructures use plan output for the same purpose.
DORA metrics
Section titled “DORA metrics”Per service, over a time range:
| Metric | Definition |
|---|---|
| Deployment frequency | Successful production deployments over time |
| Lead time for changes | Commit time to production deployment |
| Change failure rate | Share of deployments that failed or rolled back |
| Time to restore | Time from a failed deployment to the next successful one |
Available through DeployService.GetDoraMetrics and the DORA dashboard. Lead time uses the commit timestamp recorded in the release provenance.
Deployment history
Section titled “Deployment history”A timeline of every deployment with its release, strategy, approvals, gate results, verification verdicts and duration. Everything on it is also in the audit log.