Why Surveillance Networks Collapse at the Core

Posted by Jerry Tildsen on Feb 27, 2026

Why Surveillance Networks Collapse at the Core

Why Your Surveillance Network Collapses at the Core Switch (And It’s Not the Cameras)

Most surveillance problems get blamed on cameras. Or the recorder. Or the VMS.

But in larger deployments, the failure point is often the network.

Specifically: uplinks that were never sized for real aggregate video traffic.

Deployment takeaway
  • PoE port count is not network capacity. Uplinks are what carry the load.
  • Aggregate bitrate matters more than per-port speed.
  • Motion spikes can saturate uplinks even when averages look safe.
  • Unmanaged switching hides congestion and makes troubleshooting harder.

The Common Design Mistake: The 24-Port Trap

A classic surveillance build looks like this:

  • one or more 24-port PoE switches at the edge
  • a single 1G uplink from each switch back to a core switch
  • everything looks fine during light traffic or short testing windows

The issue is simple. Video traffic is constant and additive.

When enough cameras are active, the uplink becomes the bottleneck.

Reality check example
  • 24 cameras at 8 Mbps average = 192 Mbps per switch (before overhead)
  • Now add a 75% motion spike scenario = 336 Mbps per switch
  • Now run 3 access switches into a single core uplink path = ~1,008 Mbps

That is how a “normal” build quietly becomes a 1G choke point.

Access Layer vs Core Layer: What People Confuse

Many teams treat PoE switching like a checklist:

  • enough ports
  • enough watts
  • done

But PoE capacity is not network capacity.

The access switch is only one part of the path. Your core and uplinks determine whether all those camera streams can actually make it back to recording.

Symptoms of Uplink Saturation

Uplink saturation rarely shows up as a clean “network down” event. It shows up as weirdness:

  • random dropped frames or gaps in recorded footage
  • playback stuttering in VMS clients
  • slow scrubbing and delayed timeline load
  • intermittent camera disconnects that “fix themselves”
  • everything looks fine in basic monitoring because averages are misleading

When the uplink is saturated, the network starts making decisions for you by dropping packets and delaying traffic.

Quick Uplink Capacity Calculator

How many cameras feed each edge PoE switch.
Use a realistic average for your encoding settings.
Common range: 50–100% depending on environment.
How many edge switches feed your core path.

Why 10G Uplinks Are Not Enterprise Overkill

Teams often avoid 10G because it sounds “enterprise.”

In surveillance, 10G is usually just headroom. It prevents congestion during correlated spikes, firmware updates, client playback surges, and operational growth.

  • 10G uplinks protect stability when camera bitrate spikes
  • they reduce congestion impact on VMS playback
  • they provide room for future cameras without redesign

Managed vs Unmanaged Switch Risk

Unmanaged switches do not show you what is happening. They make it hard to prove the real bottleneck.

Managed switching gives you:

  • visibility into utilization and errors
  • VLAN segmentation to isolate video traffic
  • QoS and traffic control tools
  • a way to diagnose issues without guessing

How This Connects to the Full Stack

  • bitrate spikes increase storage throughput demand
  • bitrate spikes also increase uplink saturation risk
  • uplink saturation can create gaps even when storage and PoE are sized correctly

Where This Fits in a Deployment Program

Want us to validate your switching architecture?

Tell us camera count, codec, average bitrate, and how your access switches connect back to your core. We’ll help you validate uplink capacity before deployment.

Get Recommendations