Choosing a simple metrics API for Node.js health dashboards without Prometheus

Short summary

For small startups needing a basic Node.js health dashboard without Prometheus, simple metrics-and-logs APIs like Infrai work for custom counters and gauges, while Healthchecks covers scheduled-job failures. The key insight is to start with signals (healthcheck_success, queue_depth, db_ping_ms) not chart libraries, and to understand the limits: no built-in alerting, no distributed tracing, no per-user log deletion. A real-world cold-start incident shows why coarse averages can hide critical tail-latency events.

• •Simple metrics APIs suffice for basic health dashboards; keep Prometheus for alerting and tracing
• •Start with signals (counters, gauges, timings) not chart libraries
• •Coarse averages can hide critical events — pair tail-latency gauges with queue depth and dependency ping times

Generated with AI, which can make mistakes.