Business Continuity

Power outages. Severe weather. Cyberattacks. Today’s IT space can be a downtime nightmare. What’s an IT pro to do? Our business continuity topics can help you plan for disasters and get you back to business as usual if the worst happens.

How to get the most out of your UPS (article)

Getting the most value out of your UPS configuration
Like every IT manager, you work within a budget that probably isn’t enough. Limited dollars aside, you’ve still got to protect the network. So how do you get the biggest bang for your IT buck? Configuring your system based on usage and maintenance is a great way to help make every dollar count.

Configuration optimization: Centralized vs. Decentralized
It’s a conundrum every IT manager faces: Install one central enterprise-level UPS or go small with a series of decentralized, independent UPS devices. Each has budgetary benefits and drawbacks. What would work best for you? Well, it depends. Let’s first understand the basics of each.

What’s the difference?
In a decentralized UPS configuration, multiple UPSs support a handful of devices or perhaps only a single piece of equipment. Decentralized UPSs are typically “plug and play” connections that are usually less than or equal to six kVA. In a centralized UPS configuration, a single, larger UPS supports multiple devices with high kVA applications.

Pros and cons
A centralized approach requires a lot of foresight and planning as it’s fed through your electrical infrastructure and must be professionally installed. While you can expect a high capital investment, you’ll gain many monetary efficiencies over the lifespan of your system. For new construction sites, up-front UPS installation costs can often be drastically reduced by consolidating installation with many of the other electrical installation activities.

Decentralized, plug and play configurations can be inexpensive (relative to centralized), easy to install (no electrician required) and simpler to replace if needed. While many IT managers go this route, know that you’ll have to monitor and maintain each UPS. This option also a clear choice for decentralized IT, such as IDF closets, as wiring to different rooms from a centralized UPS can cause costs to skyrocket.

What’s right for you?
It’s all about what works best for your site and situation. Most centralized designs start to make sense around 10 kVA and become a more obvious preference above 20 kVA. If your critical IT gear is already running a decentralized design with around 20 kVA or less of power, continuing to utilize a decentralized design is usually the way to go. However, your strategy may change if you begin running out of electrical capacity at your panel or you are forced to make other modifications that require an electrician. In this case, a centralized approach can be both more cost effective and also allow you to plan for growth.

Finally, be sure to consider the reoccurring costs of each, including warranty, service, labor or battery replacement costs. The life span of a higher-end decentralized UPS should be between seven to 10 years, while a centralized design might last eight to 12 years. Eaton recommends replacing UPS batteries every three to four years.

Adding redundancy on a budget
For distributed network closets and server rooms, deploying fully redundant systems is an IT manager’s ultimate dream. But it comes at a price, and most IT budgets aren’t robust enough to install two UPSs for every application. So how can you achieve redundancy on a budget? Eaton’s design recommendations and accessories can help IT managers achieve the redundancy they need at a price they can afford.

Affordable redundant configuration models

Try maintenance bypass" and "Traditional installation

In most traditional power designs, one UPS supports your servers, switches and storage devices. This type of environment is prone to equipment shutdown during a power failure, UPS maintenance or UPS replacement.

Traditional installation

Most modern IT equipment has dual input cords. Deploying redundant UPSs to feed each power supply (parallel redundant) ensures that critical loads remain protected if one or more UPSs fail. Parallel redundancy offers the greatest stability for most organizations. If any UPS fails, the other systems keep protected information technology equipment (ITE) operational.

Parallel redundant installation

Single-corded IT equipment can complicate system design as a single UPS or rack PDU may represent a single point of failure. MBP can dramatically improve system reliability with marginal cost impact compared to a second UPS. Here, utility power runs through the MBP via the UPS and then supports the rest of the equipment in the rack. Should the UPS need to be replaced or serviced, the power can be switched away from the UPS without having to shut down equipment. Eaton strongly recommends an MBP for all applications, especially for installations where electricians are required for UPS replacement (typically 8kVA or above).

Maintenance bypass (MBP)

Dual feed with single corded loads
Automatic transfer switches (ATS) provide power redundancy to equipment with only one power supply. The ATS automatically transfers the power from a primary source to a secondary source if there is a problem with the primary. Once the primary source is restored, power transfers back through it.

Helpful hints for selecting an ATS:
Remotely monitor alerts, provide redundant power and keep mission critical applications running with a reliable and easy-to-use solution.

  • Graphical LCD (available on the 15- and 20-amp models) signifies which source is being output (30-amp models have a static display for identifying the source)
  • Test button can be used to manually force transfer
  • Network card supports IPv6 and provides remote monitoring capabilities via web browser or SNMP
Learn more at

Dual Feed Dual UPS
Dual Feed Single UPS

Determining power rating
A UPS’s power rating is the VA amount that it’s designed to support. Use this very basic five-step procedure to determine the approximate UPS rating your organization requires:

  • Make a list of all the equipment your UPS will protect
  • Determine how many volts and amps each device draws
  • Multiply volts by amps to determine VA figures for each device
  • Add the VA figures together
  • Multiply that sum by 1.25 to account for growth
  • The UPS purchased must have a rating equal to or greater than the final number in step five.

    Looking for a more precise calculation? Use Eaton’s UPS Selector tool.

    Virtualization keeps you connected
    One of the most important issues facing IT managers today is software/UPS integration. While virtualization plays a huge role in bridging the communication gap, many IT managers continue to run non-virtualized environments. But as graceful shutdown increasingly becomes more complicated, IT managers must take the time to learn newer integration software packages — in addition to performing their already hectic daily tasks.

    Save time with Intelligent Power Manager (IPM)

    Eaton’s vendor-agnostic software integrates with existing platforms, such as VMware, Citrix, Microsoft and Red Hat. Eaton’s IPM works seamlessly with your UPSs and rack PDUs to ensure maximum uptime, data integrity and, for some equipment, outlet level power consumption. IPM allows you to manage, monitor and control the power devices in your physical or virtual data center, without having to set aside the time to learn a new platform. Imagine it as an additional tab in your existing dashboard.

    Talking dollars and cents
    Downtime has a real impact on the bottom line. But if you’re like most IT managers, you probably find explaining the long-term cost savings that a dependable UPS system can offer difficult.

    We can help you base the conversation on real numbers. Download Eaton’s Cost Justification Worksheet to demonstrate downtime’s real costs.

    Know your power
    IT budgets are rarely robust enough to pay for everything on your wish list. Our advice: re-evaluate your needs based on power usage before you buy. This baseline will help you optimize the system and retool redundant configurations, maximizing your dollars invested while simultaneously accounting for growth.

    Was this helpful?