Your show is dying inside a box you've never opened.
Not the speakers. Not the projectors. Not the lighting rigs. The network switch, sitting in a road case or zip-tied to a truss, quietly routing every data packet your production depends on, until it can't. And when it can't, everything that reads like a tech failure, the dropped video feeds, the DMX stutter, the media server freezing mid-cue, none of that is random. That is a bottleneck you could have seen coming.
I've been building production infrastructure for over twenty years. Disney, the NBA, Dick Clark Productions, the Emmys. Shows where a glitch isn't a footnote in the debrief; it's front page. The network switch is the single most underestimated piece of gear in live event tech. Bar none.
What Actually Happens at Peak Load
Your network doesn't fail evenly. It fails at the worst possible moment, which is peak load, which is almost always showtime.
Think about what's running simultaneously when your event hits full swing. Media servers are pushing high-resolution content. Lighting consoles are firing DMX-over-Ethernet packets. Stage managers are on comms software. Your show caller is pulling video monitoring feeds. Crew tablets are on the same network. Wireless access points are daisy-chained off the same switch. All of that data is competing for bandwidth and processing priority on a device that, in most productions, got specced out once and never revisited.
Switches have a forwarding rate, measured in packets per second. They have a backplane capacity. They have buffer memory that holds packets when congestion spikes. When all three get stressed at the same moment, which is exactly when your audience is watching, packets get dropped, queued too long, or delivered out of order. The show stutters. And everyone looks at the video engineer or the lighting director, because nobody looks at the network switch.
Why Managed Switches Are Non-Negotiable on Complex Rigs
If you're running an unmanaged switch on a multi-system event rig, I've seen it hold together in smaller setups. But you have handed control of your data priority entirely to chance. Unmanaged switches treat every packet the same. Show-critical video traffic and a crew member streaming music get the same lane.
Managed switches let you configure VLANs to segment your traffic. They let you set Quality of Service rules so your lighting data and video feeds get priority handling over everything else on the network. They give you visibility at the port level, so when something starts misbehaving in tech, you can see it instead of guessing.
The top touring and corporate production companies have been running managed switch infrastructure on major shows for years. It is not new technology. It is just not universal, and the gap between who does it and who doesn't is exactly where avoidable failures live.
The Three Places the Bottleneck Actually Hides
1. The Uplink Port
You can spec a beautiful 24-port gigabit switch, and if your uplink to the next switch in the chain is a single 1Gb connection while you're pushing aggregated traffic from a dozen devices, you've built a funnel. Proper link aggregation or a 10Gb uplink between core and edge switches solves this. One cable between switches is often where the entire rig gets choked.
2. The Switch That Got Borrowed from the AV Closet
Consumer-grade and low-tier prosumer switches frequently have lower backplane capacity than their spec sheets imply. They're also cooled poorly or not at all, and thermal throttling under load in a hot equipment rack is absolutely real. I've troubleshot shows where the problem vanished the moment we swapped a hot, undersized switch for a properly rated unit. It sounds too simple. It wasn't simple at minute four of the opening number.
3. No Redundancy, Anywhere
A single switch failure should never be able to take down your entire data infrastructure. Redundant paths, whether through spanning tree protocol or a second physical switch staged and ready, mean that one device dying is an inconvenience instead of a disaster. Most show rigs I encounter don't have this. Most that had a bad night wish they did.
What Good Infrastructure Actually Looks Like
- A clearly documented network topology completed before load-in, not after something breaks
- Managed switches with VLANs separating show-critical systems from crew and guest traffic
- QoS configured to prioritize video, audio, and control data
- 10Gb uplinks between core switches on complex rigs
- Temperature-rated gear in ventilated cases, not zip-tied to whatever's nearby
- At least one spare switch on-site, pre-configured and ready to drop in
- A network monitoring view accessible to your systems tech during the show
None of this is exotic. All of it is available at reasonable cost. The barrier is almost never budget; it's the fact that network infrastructure stays invisible until it breaks, so it never gets the same planning attention that lighting or video receives.
The Real Cost of Getting This Wrong
I've seen a dropped feed cost a production a client relationship. I've seen a show that went sideways technically blamed entirely on the wrong vendor, because the actual cause, network congestion from a misconfigured switch, never got properly diagnosed. The cost isn't just the moment. It's how people retell the moment, and in this industry, the story they leave with is everything.
You can be doing everything else right. Incredible content, tight cues, a talented crew. A switch nobody properly specced can make all of it look bad.
If you want to understand what separates productions that hold together under pressure from those that don't, the services side of what I do is built around exactly that kind of infrastructure thinking. And if you're designing an event tech rig and want to work through the network before you're troubleshooting it live in front of a crowd, reach out through Ascend & Achieve and let's talk about what your rig actually needs.