At a live event, a video server failure does not wait for a convenient moment.

It can happen during a keynote presentation, halfway through a concert visual, seconds before a product reveal or while a camera feed is being displayed across a massive LED wall. And unlike a problem discovered during rehearsal, a failure during the live show is immediately visible to the audience.

That is why professional event video systems are often designed around a simple principle:

If one server is responsible for the show, another server should be ready to take over.

This is the thinking behind dual main-and-backup video servers, one of the most important forms of redundancy used in professional live event production.

A backup video server is not simply an extra computer sitting beside the main machine. In a properly designed system, the backup is configured, synchronised and tested so that it can assume responsibility for the same content and outputs if the primary system becomes unavailable.

This approach is particularly valuable for LED screens, large-format video walls, projection systems, live broadcast, corporate events, concerts, product launches, award ceremonies and other high-stakes productions where there is little tolerance for an unexpected blank screen.

Modern media-server platforms support increasingly sophisticated backup workflows. PIXERA, for example, provides full backup configurations in which a mirrored backup system can follow the main system’s playback, with the output switched to the backup in the event of failure.

The concept is straightforward:

Main runs the show. Backup is ready to run the show.

But making that switch reliable requires much more than installing two machines.

What Is a Video Server in Live Event Production?

Before discussing redundancy, it is useful to understand what a video server actually does.

A professional media server is a high-performance computer system designed to manage, process and output video content in real time.

Depending on the production, a video server may handle:

• Playback of pre-rendered video
• Motion graphics
• Still images
• Live camera inputs
• Multiple outputs
• LED processor feeds
• Projection mapping
• Layer compositing
• Real-time effects
• Screen configuration
• Content switching
• Cue-based playback
• Timecode
• Show control
• Synchronisation with other systems

For a large LED wall, the media server may sit between the creative content and the display-processing infrastructure.

A simplified workflow could look like:

Content → Media Server → Video Processing → LED Processor → LED Wall

If the media server fails, everything downstream may continue functioning perfectly—but there may be no usable video signal reaching the screen.

That is why the media server is considered a critical point in the video chain.

Why One Video Server Is Not Always Enough

Imagine a corporate conference where the main stage uses a 20-metre LED wall.

The screen is displaying:

• Speaker presentations
• Animated backgrounds
• Sponsor logos
• Event branding
• Camera IMAG
• Transition graphics
• Opening videos
• Closing videos

The event has taken months to organise.

The venue is full.

The keynote speaker is on stage.

Then the media server crashes.

The LED processor is still powered.

The LED cabinets are still powered.

The cables are still connected.

The content files still exist.

But the screen has lost its video source.

Technicians may be able to restart the server, but restarting a production system in the middle of a live show is not always practical. Depending on the failure, the system may require troubleshooting, software recovery, hardware replacement or a complete reboot.

Even a short interruption can be highly visible.

This is where redundancy changes the risk profile.

Instead of asking:

“How quickly can we repair the failed server?”

the production team can ask:

“Can we move the show to the backup system?”

That is a much better position to be in during a live event.

What Does “Main & Backup” Actually Mean?

A typical redundant configuration consists of two systems:

Main server

The main server is the primary system responsible for running the show.

It handles the live playback and normally sends the active outputs to the video-processing system.

Backup server

The backup server is configured to mirror the main system as closely as practical.

Depending on the workflow, it may:

• Have the same project
• Have the same media
• Use matching output configurations
• Follow the same playback timeline
• Receive the same control commands
• Maintain synchronisation with the main server
• Remain ready for immediate or rapid takeover

The backup does not necessarily mean the system is doing nothing.

In many professional workflows, the backup system is actively following the main system so that it is already in the correct state if the primary system fails.

PIXERA’s documentation describes a full-backup configuration where the backup system mirrors the live system and follows playback updates, allowing the output to be switched to the backup system when required.

That distinction is important.

A spare computer is not automatically a backup system.

A real backup system needs to be operationally prepared.

Cold Backup, Warm Backup and Hot Backup

Not all backup strategies provide the same level of protection.

Cold backup

A cold backup is essentially a spare system that is powered down or not actively participating in the show.

If the main server fails, technicians start the backup and restore the necessary project.

This can be useful for lower-risk productions, but the recovery time can be significantly longer.

Warm backup

A warm backup is powered and prepared but may not be continuously outputting or synchronised with the main system.

It can be brought online more quickly than a cold spare.

However, the system may still require some manual intervention before it is ready to take over.

Hot backup

A hot backup is designed to be immediately or rapidly available.

The backup system is typically running alongside the main system and maintaining the appropriate show state.

If the main server fails, the output can be switched to the backup.

This is generally the preferred approach for productions where interruption must be minimised.

However, the exact behaviour depends on the media-server platform, switching architecture, networking and event design.

The Most Important Word Is “Synchronised”

Having identical content on two servers is not enough.

Imagine the main server is currently playing:

00:03:27 of a 10-minute video.

The backup server has the same video—but it is sitting at:

00:00:00.

If the main fails and the backup starts from the beginning, the audience will immediately notice.

A properly designed backup workflow therefore aims to keep the two systems aligned.

Depending on the system, synchronisation can involve:

• Timeline position
• Playback commands
• Cue states
• Content updates
• Show control
• Timecode
• Input routing
• Output configuration
• Project data
• Media assets

PIXERA’s multi-user functionality, for example, can keep systems updated so that a backup system is prepared to replace a failed system. Its documentation also describes distributing content to required live systems, including backup systems.

This is one of the major differences between a professional redundant setup and simply keeping a second laptop nearby.

How Does the System Switch From Main to Backup?

The actual switching mechanism depends on the architecture.

A simplified example is:

Main Server → Video Matrix / Switcher → LED Processor → LED Wall

and

Backup Server → Video Matrix / Switcher → LED Processor → LED Wall

The output device or matrix determines which server is currently feeding the display.

If the main system fails, the operator can switch the video path to the backup.

Some media-server systems can integrate backup triggers directly into the workflow. PIXERA documentation describes backup scenarios where a matrix switcher can be triggered to change from the main output to the backup output.

This means redundancy is not necessarily limited to the servers themselves.

It can extend throughout the signal chain.

True Redundancy Means Eliminating Single Points of Failure

This is where professional system design becomes especially important.

Suppose you have:

• Two media servers
• One power socket
• One network switch
• One video switcher
• One output cable

You technically have two servers.

But the system is still vulnerable.

If both servers depend on the same failed component, the second server cannot save the production.

This is known as a single point of failure.

A stronger redundant architecture considers the entire video chain.

That can include redundancy for:

• Media servers
• Power
• Network paths
• Video outputs
• Matrix switching
• Signal distribution
• Processing
• Control systems
• Timing references
• Storage
• Critical network infrastructure

The exact architecture should be proportional to the event’s requirements and risk.

Power Redundancy Matters Too

A backup server that shares the same vulnerable power source as the main server is not fully independent.

Professional production environments may use separate power feeds, UPS systems or appropriately designed power distribution to reduce the likelihood that a single electrical problem takes down both systems.

This is particularly relevant at temporary event sites, where production systems may be operating alongside lighting, audio, catering, venue equipment and other electrical loads.

Power planning should therefore be part of the redundancy discussion from the beginning.

Network Redundancy Is Becoming More Important

As live production systems increasingly use IP-based infrastructure, network design is becoming a major part of redundancy planning.

Modern standards such as SMPTE ST 2110 define professional media transport over managed IP networks and address the carriage, synchronisation and description of separate video, audio and ancillary-data streams.

For large productions, this can enable highly flexible signal distribution—but it also means network infrastructure becomes part of the production-critical system.

A redundant IP workflow may therefore consider:

• Multiple network paths
• Redundant switches
• Separate network interfaces
• Appropriate bandwidth
• Timing and synchronisation
• Network monitoring
• Failover behaviour

SMPTE has specifically highlighted the use of dual IP streams for redundancy in professional media facilities.

The lesson is simple:

As the video system becomes more networked, network reliability becomes part of video reliability.

Synchronisation and Timing

A backup server must know where the show is.

That is why timing and synchronisation are so important.

Depending on the production, systems may use technologies such as:

• Timecode
• Genlock
• PTP
• Network-based synchronisation
• Media-server software synchronisation
• Show-control protocols

SMPTE ST 2110 systems use a common reference clock and timestamping mechanisms to maintain timing relationships between media streams.

In large-scale IP production environments, Precision Time Protocol (PTP) can become an important part of the timing architecture.

This is particularly important when multiple systems need to operate together rather than simply producing independent video outputs.

Why Backup Servers Are Critical for LED Walls

Large LED screens are unforgiving.

A small screen problem might go unnoticed.

A 2-metre display going black is one thing.

A 30-metre stage-wide LED wall going black is something else entirely.

LED walls are increasingly used as major visual elements in:

• Concerts
• Award shows
• Corporate conferences
• Product launches
• Exhibitions
• Festivals
• Government events
• Fashion shows
• Sports productions
• Broadcast environments
• Hybrid events

The larger the display, the more noticeable a video failure becomes.

This is one reason large-scale LED deployments often justify stronger redundancy planning.

A recent SMPTE case study on the Eurovision Song Contest’s transition to an ST 2110-based infrastructure described a 100 Gb dual-redundant backbone supporting a large event environment and specifically highlighted redundancy and network distribution for large LED display systems.

That illustrates an important principle:

The bigger and more technically complex the production, the more important it becomes to design redundancy into the architecture rather than adding it as an afterthought.

Backup Servers Also Protect Against Human Error

Redundancy is usually discussed in terms of hardware failure.

But live events also involve people.

Operators can accidentally:

• Close software
• Load the wrong project
• Change an output
• Disconnect a signal
• Trigger the wrong cue
• Modify a setting
• Overload a system

A properly designed backup architecture can provide another layer of protection.

It does not eliminate human error, but it can reduce the impact of certain mistakes.

This is another reason why operator training and show procedures should be treated as part of redundancy planning.

Testing the Backup Is Just as Important as Installing It

One of the biggest mistakes in redundancy planning is assuming that the backup works because it exists.

It needs to be tested.

A proper pre-show test should simulate realistic failure scenarios.

For example:

Test 1: Main server failure

Stop the primary server and verify that the backup takes over correctly.

Test 2: Output failure

Disconnect or disable the relevant main output path and verify that the backup path works.

Test 3: Network failure

Where appropriate, test the network path used by the production without creating unnecessary risk to the live environment.

Test 4: Control failure

Verify what happens if the main control interface or communication path becomes unavailable.

Test 5: Power interruption

Where safely and appropriately designed, test the intended power-failover behaviour.

The goal is not simply to prove that the backup server turns on.

The goal is to prove that the show can continue.

Why Manual Backup Can Sometimes Be Better Than Automatic Backup

It might seem that automatic failover is always the best option.

Not necessarily.

In a live environment, false triggers can create their own problems.

Some media-server systems therefore recommend different approaches depending on the application.

PIXERA’s documentation, for example, recommends manual-triggered backup scenarios for live systems while noting that automatic backup settings are more appropriate for fixed installations where a technical operator may not be present.

This highlights an important principle:

Automation should be designed around the production—not used simply because it is technically possible.

A live event with an experienced video operator may benefit from controlled manual failover.

A permanent installation with no technician on site may require automated failover.

The correct approach depends on the risk profile.

Redundancy Does Not Mean “Identical Everything” Without a Plan

It is tempting to assume that the safest approach is to purchase two identical servers and stop there.

But redundancy requires more than matching specifications.

The systems should be aligned in areas such as:

• Software version
• Project version
• Media files
• Output configuration
• Graphics drivers
• Hardware configuration
• Network settings
• Control mappings
• Calibration data where applicable
• Display routing
• Synchronisation settings

A software update applied to the main server but not the backup can create a dangerous situation.

The same applies to content.

If the main system receives a last-minute version of a presentation but the backup still contains yesterday’s file, the backup is no longer an accurate mirror.

Content Management Is Part of Redundancy

This is an often-overlooked issue.

Suppose the main and backup servers are perfectly synchronised—but the latest 4K video file only exists on the main machine.

If the main fails, the backup cannot play content it does not have.

That is why professional workflows include content distribution and verification.

PIXERA’s resource-distribution documentation, for example, includes an option to distribute content to systems that require it, including backup systems.

Before showtime, teams should verify:

• All media exists on both systems.
• File paths are correct.
• Required codecs are supported.
• Content versions match.
• Playback settings match.
• External media dependencies are available.
• The backup can actually decode and play the content.

A backup without the latest content is only partially redundant.

What Happens During a Real Failure?

A simplified live-event sequence might look like this:

1. The main server encounters a fault.

The operator notices a loss of output, software failure or system error.

2. The operator confirms the issue.

The team determines whether the problem is actually the main server or another component.

3. Backup is activated.

The backup server is selected as the active source.

4. Video routing changes.

The matrix, switcher or appropriate signal path moves the display from Main to Backup.

5. Backup continues playback.

Because the backup was synchronised, it can continue from the appropriate point in the show.

6. The audience sees minimal disruption.

The entire purpose of the system is to keep the visible impact as small as possible.

This is what real-time redundancy is designed to achieve.

A Practical Main & Backup Checklist

For an event using redundant media servers, production teams should consider checking:

• Main and backup servers have compatible hardware.
• Software versions have been verified.
• The latest show project exists on both systems.
• All media assets are available on both systems.
• Output configurations match.
• LED/projection output routing has been verified.
• Main and backup systems are synchronised.
• Timing references have been checked.
• Network connectivity has been tested.
• Backup power arrangements have been reviewed.
• Video switching has been tested.
• Failover procedures are documented.
• Operators know who is responsible for triggering failover.
• A realistic failure test has been completed before doors open.
• Backup content has been checked after the final content update.
• Any last-minute show changes have been mirrored to the backup.
• Spare hardware and appropriate technical support are available where required.

Redundancy Is About Risk Management

No technical system can guarantee that a live event will experience zero failures.

The purpose of redundancy is different.

It is about reducing the consequences of failure.

If a single server failure can stop an entire presentation, then that server represents a significant production risk.

If a second synchronised server can take over, the same failure becomes a manageable technical incident rather than a potentially show-stopping event.

This is the fundamental philosophy behind redundancy.

You are not paying for a second server because you expect the first one to fail.

You are deploying the second server because you cannot afford to discover what happens if it does.

Where Dual Video Servers Make the Most Sense

Not every event requires the same level of redundancy.

However, dual main-and-backup media servers are particularly valuable for:

Large corporate events

Important keynote presentations, executive speeches and product launches often depend heavily on video content.

Concerts and live entertainment

A failed visual during a headline performance can be immediately noticeable to thousands of people.

Broadcast and livestream productions

Camera-facing LED walls and graphics systems must remain reliable throughout transmission.

Award ceremonies

High-value branded content and stage visuals often run continuously throughout the show.

Product launches

When the LED wall is part of the reveal itself, there is little room for technical failure.

Large-format LED installations

The larger the screen, the more visible a complete video failure becomes.

High-profile government and institutional events

Reliability and professionalism are particularly important when there is significant public visibility.

The Future of Redundant Live Video Systems

Live production is increasingly moving toward software-defined, networked and distributed systems.

SMPTE describes ST 2110 as a key standard for professional media over managed IP networks, supporting real-time production and playout while separating video, audio and ancillary-data streams.

This evolution creates new possibilities for redundancy.

Instead of simply thinking about:

Server A + Server B

production engineers increasingly need to consider:

Primary + Backup + Network + Timing + Processing + Routing + Power + Monitoring

That creates a more resilient production architecture.

At the same time, redundancy must remain practical. More equipment means more configuration, more monitoring and more opportunities for setup errors if the system is poorly managed.

The goal is therefore not to add equipment for the sake of having more equipment.

The goal is to build a controlled failure strategy.

In conclusion: The Backup Is Part of the Show

A successful live event often makes complex technology look effortless.

The audience sees the LED wall.

They see the graphics.

They see the videos.

They see the camera feeds.

They see the stage.

What they do not see is the technical planning happening behind the scenes to make sure those visuals remain on screen.

A dual main-and-backup video server setup is one part of that planning.

It provides a second path for critical video playback, helping production teams respond when a primary server, output or associated system encounters a problem.

But true redundancy is not simply about having two computers. It involves synchronisation, content management, signal routing, timing, power, networking, operator procedures and—most importantly—real-world testing.

For major LED and live-event productions, redundancy is therefore best viewed as an investment in continuity and confidence.

At DOREMi Events, reliable event technology is about more than getting a screen to light up. It is about designing and managing the technical workflow behind the screen so that the production remains dependable when the pressure is highest.

Because when thousands of people are watching, the best backup system is the one the audience never knows was needed.

Main runs the show. Backup protects the show. And proper redundancy makes both work as one production system.

Explore More on Network Redundancy

Reliable event systems depend on eliminating single points of failure and maintaining continuity when equipment, connections or network paths fail. For further insights into network redundancy, failover mechanisms and high-availability network design, explore SPON Communications’ Network Redundancy: Definition, Types, Failover & Why It Matters for Modern Networks.