Skip to content
In production

Energy

Power supplies, PDUs, UPS and energy equipment watched over SNMP, in the same place as the cameras and the assets.

Why it matters

When a site goes down, the first question is whether it was the network or the power, and there's almost never a way to know: the energy gear lives in its own portal, with its own login, and nobody goes in until something has already happened. The consequence is that a «camera down» alarm and a power cut look exactly the same from the console.

How it works

  1. 01

    Purpose-built SNMP driver for v1, v2c and v3, with per-model profiles: the OID map belongs to the device, not to a generic table that almost gets it right.

  2. 02

    Beyond polling, trap reception: the device reports for itself when something happens, without waiting for the next round. Measured at 445 ms from the event to the alarm existing.

  3. 03

    Thresholds are set on the DEVICE, with hysteresis — not in a rule written elsewhere that nobody remembers to update.

  4. 04

    Each output is linked to what it feeds, so the alarm says what went dark, not just that a circuit opened.

  5. 05

    Fleet view: every energy device across every site on one screen, with the reason when one doesn't answer.

Energy
12,4 kW · fase B

The difference

The alarm says WHAT went dark

A PDU portal tells you circuit 3 opened. At three in the morning that helps nobody: someone has to walk to the rack, read the label and guess what hung off it. When each outlet is linked to the equipment it feeds, the alarm names the machine — and that is the whole difference between a datum and a decision.

PDU · 8 outletsSNMP v3 · per-model profileS1North entrance cameraS2Main rack switchS3Building A recorderS4Substation thermal cameraSIN VÍNCULO«Circuit 3 opened.»CON VÍNCULO«No power: Building A recorder.»

The threshold is set by the EQUIPMENT, with hysteresis, not by a rule written elsewhere that nobody remembers to update. And besides polling there is trap reception: the device reports by itself, measured at 445 ms from the event to the alarm existing.

In practice

Why that site went down

When a site stops reporting, the console cross-checks power state against camera state. If the supply lost input and the cameras dropped after it, the diagnosis is done — and nobody has to drive out to look.

Capabilities

  • SNMP v1, v2c and v3 with per-model profiles
  • Traps received, not just polling · 445 ms measured
  • Set-points on the device, with hysteresis
  • Output → powered equipment link
  • Multi-site fleet on one screen
  • Correlation with camera outages
  • Onboarding from the admin console

What it does not do yet

  • The module is tested against generic SNMP equipment and a DIN-rail PDU profile. Every new model comes in with its profile; it gets done, but it isn't automatic.

See, understand and act in time.