VMS, CMS and the problem of knowing your cameras are actually working

Most organisations discover a camera has been offline for weeks at exactly the wrong moment — when they go looking for footage of an incident. The uncomfortable truth about physical security estates is that recording and knowing you are recording are different problems, and most deployments only solve the first.

What a VMS does

A video management system is the software layer that controls how video is captured, stored, viewed and analysed. It aggregates feeds from cameras regardless of manufacturer, handles retention and playback, manages user permissions, and increasingly runs analytics over the stream. It is the thing operators actually sit in front of.

What a VMS is not, in most deployments, is an uptime monitoring system. It shows you the cameras that are streaming. It is generally less good at loudly telling you about the ones that are not, across dozens of sites, before anyone needs the footage.

Where camera monitoring fits

Camera monitoring — sometimes deployed as a camera monitoring system, or CMS, alongside the VMS — is concerned with health rather than viewing. The questions it answers are operational: which cameras are offline right now, which NVRs have failed disks, where has recording silently stopped, which sites have degraded retention, and how long has any of it been true.

• Live online/offline state for every camera across every site
• NVR and storage health, including disk and retention shortfalls
• Recording continuity — not just whether a feed is up, but whether it is being written
• Stream quality degradation, not only hard failures
• Uptime reporting suitable for an audit or an SLA conversation

Why multi-vendor is the hard part

Real estates accumulate brands. Hikvision and Dahua dominate by volume, CP Plus is common across Indian deployments, and Axis, Bosch and Hanwha appear wherever specification mattered more than price. Each exposes status differently through its own SDK and management tools.

Two standards make vendor-agnostic monitoring practical. ONVIF provides a common interface for device discovery and control, and RTSP provides a common way to pull the stream itself. A platform built on those can assess health across brands without a separate integration per manufacturer — which is what makes one dashboard across a mixed estate achievable rather than aspirational.

The correlation nobody does

Camera health is usually treated as a facilities problem, which is why it sits apart from IT monitoring. That separation costs you. A camera going offline can be a failed device, a PoE switch port, an uplink saturating, or someone unplugging it deliberately — and those are network, infrastructure and security events respectively. When camera telemetry lands on the same plane as network and security signals, the pattern is visible. When it does not, three teams each see a fragment.