Live demo

A real Kubernetes cluster.
Deliberately broken every few minutes.

Reflex Runtime is watching demo-payments-api on a real single-node k3s cluster. Every 4–9 minutes the service starts returning 5xx. Reflex detects the SLO breach, posts a Slack approval, and restarts the deployment. The state below is live — the events, the counter, the countdown.

Get started freeRead the docs
Service status
Healthy
payments-api
Next breach in
1:58
Randomized 4–9 min
Total remediations
1,114
Last: 4m ago

What this looks like for your team

Reflex has caught 1,114 breaches on this service.

Avg breach → recovery
81s
Overnight (12am–6am ET)
25%
Est. on-call hours saved
74.3

Hours-saved uses a conservative 4 min/page estimate. Real on-call interrupts — page → diagnose → fix → cool down — typically run 15–30 min.

Incident log

Live feed of breach and recovery events from the demo cluster.

  • Recovered · restart (dry-run) · 60s4m ago
  • Breach detected · error rate 70%6m ago
  • Recovered · restart (dry-run) · 60s12m ago
  • Breach detected · error rate 67%14m ago
  • Recovered · restart (dry-run) · 60s18m ago
  • Breach detected · error rate 87%20m ago
  • Recovered · restart (dry-run) · 121s26m ago
  • Breach detected · error rate 68%28m ago
  • Recovered · restart (dry-run) · 61s32m ago
  • Breach detected · error rate 86%34m ago

What happens during a breach

The exact Slack message your on-call would receive.

#reflex-demo·Slack
R
ReflexAPPjust now
SLO breach detected
demo-payments-api error rate at 70% — exceeds 5% threshold over 5m.
Pattern: error_spike
Recommended action: restart deployment/demo-payments-api
Trust mode: dry-run

Breach cycles randomized 4–9 minutes · All remediations are real · This is a real Kubernetes cluster running Reflex Runtime.