Data Center Power Redundancy: A+B Feeds, ATS and UPS

Moving business storage into a data center is about more than obtaining rack space and a fast internet connection. One of the biggest advantages of professional colocation is access to power infrastructure designed to keep critical equipment operating when individual electrical components or utility services fail.

Modern facilities can combine A+B power feed server hosting, uninterruptible power supplies, generators, Automatic Transfer Switches, and servers with redundant power supplies to reduce single points of failure.

For organizations hosting Synology systems, backup appliances, virtualization servers, and other critical infrastructure, understanding data center power redundancy helps explain why professional colocation can provide greater resilience than a typical office server closet.

Why Power Redundancy Matters

Servers depend on continuous electrical power.

Even a brief interruption can affect:

• File services
• Backup jobs
• Virtual machines
• Databases
• Remote employee access
• Replication
• Hosted applications
• Surveillance systems

Unexpected shutdowns can also interrupt storage operations and create additional recovery work.

A well-designed data center therefore assumes that electrical components can fail and provides alternative power paths to keep equipment operating.

What Are A+B Power Feeds?

A+B power architecture provides two separate power feeds for supported equipment.

They are commonly referred to as:

A Feed: Primary independent power path

B Feed: Secondary independent power path

Rather than connecting every component to a single electrical source, appropriately designed equipment can receive power from both feeds.

If one power path becomes unavailable, the other can continue supplying power.

The amount of independence between A and B paths depends on the actual facility design, so businesses should verify the architecture offered by their colocation provider.

Dual Power Supplies at the Server

A+B feeds are especially useful when the server has redundant power supply units.

A server with two PSUs can typically connect:

PSU 1 → A power feed

PSU 2 → B power feed

If one PSU, PDU, circuit, or upstream power path fails, the remaining power supply can continue supporting the server, provided the system is designed and loaded appropriately.

This helps eliminate the server’s individual power supply as a single point of failure.

Enterprise rackmount systems frequently support redundant PSUs for this reason.

What About Single-Power-Supply Equipment?

Not every NAS, network appliance, or server has dual power supplies.

A single-cord device cannot simply connect simultaneously to independent A and B feeds.

This is where an Automatic Transfer Switch (ATS) can become useful.

An ATS can accept two appropriate power sources and provide power to equipment through a single output path. If the preferred source fails, the switch transfers the load to the alternate source.

The exact switching architecture and supported equipment should be reviewed carefully before deployment.

Understanding Automatic Transfer Switches

An Automatic Transfer Switch monitors available power sources and transfers a connected load when the active source becomes unavailable.

Within a colocation environment, ATS technology can help provide redundancy for equipment that does not natively support dual power inputs.

For example:

A Feed + B Feed → ATS → Single-Power-Supply Device

This does not turn the device itself into a dual-PSU server, but it can reduce dependence on a single upstream power source.

ATS equipment should be appropriately sized for the connected load and deployed according to the facility’s electrical design.

Where UPS Systems Fit

An Uninterruptible Power Supply provides temporary power when the normal electrical source becomes unavailable or unstable.

In a professional data center, UPS infrastructure can bridge the period between utility power loss and activation of longer-duration backup generation.

UPS systems can also help protect equipment from certain electrical disturbances.

This is important because backup generators do not necessarily assume the load instantaneously when utility power fails.

The UPS provides continuity during that transition.

What Happens During a Utility Power Failure?

A properly engineered power architecture uses multiple layers rather than depending on one backup device.

A simplified sequence might look like this:

1. Utility power becomes unavailable.
2. UPS infrastructure continues supporting critical equipment.
3. Generator systems start.
4. Electrical switching transfers the required load.
5. Generators provide longer-duration backup power.
6. Utility power is restored.
7. The facility transitions back according to its operating procedures.

Actual data center electrical architectures can be considerably more complex, particularly in facilities with redundant UPS systems, multiple generators, independent distribution paths, and maintenance bypass capabilities.

Why Diesel Generators Are Important

UPS batteries are generally designed to provide transitional or limited-duration backup power, not indefinitely operate an entire data center.

Generators provide longer-duration emergency power.

Facilities may incorporate:

• Multiple generators
• On-site fuel
• Fuel-management procedures
• Generator monitoring
• Regular testing
• Redundant generation capacity

Generator infrastructure allows critical systems to remain operational during extended utility outages, subject to the facility’s design, operating conditions, and fuel availability.

Power Distribution Inside the Rack

Redundancy does not stop at the facility electrical room.

Power must travel through the infrastructure to the individual server.

A redundant path can involve:

Utility/Generator → UPS → Distribution Infrastructure → PDU → Server PSU

When two independent paths are provided, the server can potentially continue operating when a component on one side fails.

Connecting both redundant server power supplies to the same PDU would undermine much of this benefit.

Proper rack design therefore matters.

Avoid Overloading One Power Feed

Having two power feeds does not automatically provide useful redundancy if the load is incorrectly distributed.

Administrators need to understand what happens if either feed disappears.

If losing the A feed forces the B feed to carry more equipment than its supported capacity, the supposed backup path may not provide the expected resilience.

Power planning should therefore account for normal operation and failure scenarios.

Power Redundancy and Synology Colocation

Businesses increasingly colocate Synology systems because storage infrastructure may need to remain available around the clock.

A colocated Synology NAS can support:

• Business backups
• Microsoft 365 protection
• Google Workspace backups
• Private cloud storage
• File synchronization
• Disaster recovery
• Hosted applications
• Remote collaboration

Rackmount Synology systems with redundant power capabilities can benefit from appropriately designed A+B infrastructure.

Single-power-input systems may require a different redundancy strategy.

Data Center vs. Office Server Closet

Many small and medium businesses operate servers from office environments originally designed for employees rather than continuous IT infrastructure.

An office may depend on:

• One utility feed
• One electrical panel
• Consumer-grade UPS devices
• Standard building cooling
• One internet connection
• Limited generator availability

Building equivalent redundancy internally can become expensive.

Professional colocation allows organizations to place their own hardware inside infrastructure already designed around power, cooling, physical security, and connectivity requirements.

Understanding Tier 3 Facility Power Uptime

Organizations researching tier 3 facility power uptime should understand that data center tier terminology refers to defined infrastructure characteristics, not simply whether a facility owns a generator.

Tier-related designs can involve redundancy and maintainability requirements across critical infrastructure.

Businesses should verify the specific certification, topology, redundancy, service-level commitments, and operating procedures of any facility they are evaluating rather than assuming that marketing language guarantees zero downtime.

Even highly redundant infrastructure cannot make an absolute promise that equipment will never experience an outage.

Test the Complete Redundancy Strategy

Redundant equipment provides the greatest value when the architecture is documented and periodically tested. Design resilient colocation with expert Synology consulting.

Organizations should understand:

• Which PSU connects to which feed
• PDU capacity
• ATS configuration
• Expected load during feed failure
• UPS protection
• Generator coverage
• Network redundancy
• Recovery procedures

Power should also be considered alongside storage, networking, backup, and disaster recovery planning.

A server that remains powered but loses all network connectivity is still unavailable to users.

Building More Resilient Synology Infrastructure

Reliable colocation is created through layers. A+B power feeds protect against individual electrical-path failures, dual PSUs provide server-level redundancy, ATS equipment can support appropriate single-cord devices, UPS systems bridge short-term interruptions, and generators provide longer-duration emergency power.

For organizations that already own Synology hardware but want to move it beyond the limitations of an office server room, Epis Technology’s Bring Your Own Synology (BYOS) Colocation solutions provide an option for placing compatible infrastructure in a professionally managed hosting environment.

About Epis Technology

Epis Technology helps organizations design, deploy, host, and protect business-critical Synology infrastructure. Services include BYOS colocation, Synology consultation, storage architecture, backup implementation, network planning, disaster recovery, cybersecurity, and ongoing managed support. Epis Technology helps businesses move critical storage beyond traditional office server closets and build resilient environments around secure data center infrastructure, reliable connectivity, and layered business continuity planning.