Add Runway deployment pipeline metrics
Context
Right now, Runway does not emit SLIs for deployments: https://gitlab.com/gitlab-com/runbooks/-/tree/master/docs/runway?ref_type=heads#performance.
When a deployment failure occurs, it must be both manually detected (eg https://gitlab.com/gitlab-com/gl-infra/platform/runway/deployments/ai-gateway/-/pipelines?page=1&scope=all&status=failed) and manually escalated (eg https://gitlab.slack.com/archives/C052QHHFNH0/p1704770111694499).
In &1206 (comment 1782283120), we decided that we must lower our risk tolerance for deployment pipeline failures.
What
Update runwayctl
to emit custom SLIs for deployment frequency, deployment failure, and deployment duration.
Why
To provide a stable platform, Runway service owners and operators must have confidence that changes are consistently and successfully deployed as expected instead of silently failing.