Skip to content

Docs

Team and Enterprise

Risk radar

Risk radar surfaces latent risks across your fleet before they page someone — capacity pressure, unhealthy workloads, and access that needs attention. It's read-only: it reads live state and recommends, it changes nothing.

What it watches

  • Disk & memory pressure — a server whose latest fresh metric is high (medium as it climbs, high near full). Offline servers with stale metrics never raise a false alarm.
  • Unhealthy resources — workloads in a failing state (crash-looping, errored, not-ready) across Kubernetes and Docker.
  • Aging access — just-in-time access requests sitting unapproved too long.
  • Recent break-glass — emergency access used recently, so it gets a second look.

Reading the board

Open Risk radar from the sidebar. Findings are ordered high → medium → low, each naming the subject (a server, a count) and what to do. An empty board means capacity, workloads and access all look healthy right now. Refresh to re-scan.

Risk radar is advisory — it points you at what's worth a look. Acting on a finding (freeing disk, approving or revoking access, fixing a workload) goes through the usual governed paths.

Permissions

server.read

The board needs server.read and is hidden until an administrator enables it for your workspace.