Why monitor your infrastructure

Why monitor

Without monitoring, you find out about problems when users complain — too late. Good monitoring lets you:

• Catch problems early, before they hit the service
• Diagnose faster with historical data instead of guesswork
• Plan capacity by watching growth trends
• Prove SLA compliance with verifiable metrics
• Find bottlenecks before they become outages

Key metrics

Network

Devices

Monitoring protocols

• SNMP: the main protocol for reading metrics from network gear. Works by polling — the monitor queries each device on a schedule (typically every 60-300 seconds). Universal vendor support. See [[crearack—redes-infra—conceptos-snmp]].
• ICMP (ping): the simplest availability check. Send a packet, wait for the reply. Fast and needs no configuration on the device, but only tells you “reachable or not”.
• HTTP/HTTPS checks: verify a web service responds correctly — status code, response time, expected content.
• Syslog: devices push log messages to a central server in real time. Complements polling with event detail.

Alerts without the noise

A good alert strategy avoids both false alarms (alert fatigue) and silence during real incidents:

1. Set realistic thresholds based on your normal baseline, not arbitrary numbers.
2. Require persistence: don’t alert on a momentary spike — e.g. CPU > 90% for 5 minutes.
3. Escalate progressively: warning to the on-duty tech, critical to the manager.
4. Avoid cascades: if a router dies, don’t fire alerts for everything behind it.
5. Review regularly: delete alerts everyone ignores, tune the rest.

How CreaRack covers this

CreaRack splits infrastructure monitoring into focused modules, all feeding the same alert system:

• [[crearack—monitoring—que-es-observatory]] — Network Observatory, the core: SNMP polling, ping and HTTP checks on your network devices, with real-time dashboards and historical charts. Metrics are stored in a time-series database with 180-day retention, so you can answer both “what is happening now” and “what changed last quarter”.
• [[crearack—monitoring—que-es-ups]] — UPS Monitor: battery charge, load and runtime of your uninterruptible power supplies. This is how you cover the power layer described in [[crearack—redes-infra—alimentacion]].
• [[crearack—monitoring—que-es-wireless]] — Wireless Monitor: WiFi access points and their controllers — radios, SSIDs, connected clients.
• [[crearack—monitoring—alertas]] — threshold alerts across all of the above: define the metric, the threshold and the persistence, and CreaRack notifies you when it trips.

Suggested polling intervals

Shorter intervals give finer detail but generate more traffic and stored data. Start conservative and tighten only where you need it. In CreaRack, this is exactly what the per-device Polling cadence (15 s to 5 min) controls — see [[crearack—network—device-page]]; availability pings automatically stay at 30 s or faster regardless of the cadence you pick.

Related

• [[crearack—monitoring—que-es-observatory]] — Network Observatory overview
• [[crearack—monitoring—que-es-ups]] — UPS Monitor overview
• [[crearack—monitoring—que-es-wireless]] — Wireless Monitor overview
• [[crearack—monitoring—alertas]] — creating threshold alerts
• [[crearack—monitoring—ping-http]] — ICMP and HTTP checks
• [[crearack—monitoring—configurar-snmp]] — set up SNMP polling
• [[crearack—redes-infra—conceptos-snmp]] — SNMP in depth
• [[crearack—network—device-page]] — the polling cadence dial

Véase también

• [[crearack—monitoring—que-es-observatory]]
• [[crearack—monitoring—que-es-ups]]
• [[crearack—monitoring—que-es-wireless]]
• [[crearack—monitoring—alertas]]
• [[crearack—monitoring—ping-http]]
• [[crearack—monitoring—configurar-snmp]]
• [[crearack—redes-infra—conceptos-snmp]]