Migrating your data center is a complex process that requires a high degree of coordination with little tolerance for failure.  The goals of any move are to:

• Minimize or eliminate downtime during the move
• Maintain or enhance the performance, disaster recovery and business continuity support of the data center

There are three key steps to planning a successful data center migration:

1. Complete an audit of your current data center

The first step to a successful migration is to complete a top to bottom assessment of what’s in your current data center.

Such an assessment should start with an evaluation of the applications your data center supports.  Make sure you are thoroughly up to speed on:

• Application owners
• Planned upgrades or changes
• Allowable downtime
• Server, network, storage and operating system requirements

Next you’ll need to document your existing infrastructure, including:

• Equipment
• Power specifications
• Cooling requirements
• Connectivity diagrams
• Rack layout

The biggest mistake data center managers make in planning a migration is attempting to set schedules and budgets before completing a thorough data center audit.

2. Map existing functionality to new data center

Moving your data center presents a logical opportunity to improve the efficiency of your data center by replacing outdated hardware, upgrading software, and otherwise improving the technology in your data center.

A migration may also allow you to  reduce energy consumption and improve the reliability of your equipment through a properly designed infrastructure plan.  Keep in mind that cooling, power and IT infrastructure requirements may change as you upgrade your data center’s hardware or modify how it is used.

As you plan your new data center:

• Consider virtualization or equipment upgrades.
• Consider security, storage and monitoring requirements.
• Design you new data center architecture
• Design power, cooling and IT infrastructure to support the data center

While the IT team is responsible for what goes in the data center, having a partner that’s experienced in the design and installation of power, cooling and IT infrastructure is key to a successful migration.

3. Plan your move

Any data center migration consists of a long list of specific tasks that must be completed at both ends of the move.  There are a number of specialized vendors and service providers that must be coordinated to ensure a smooth relocation.

There are few companies who can truly serve as a single point of contact for your data center manager during this process.  You will need a partner who can bridge the gap between facilities and IT and has experience managing multiple vendors through data center consolidations and relocations.

Which Cloud Exactly?

There are three options to consider when moving to the cloud:

3rd Party Cloud

One option is to move all or part of your data center functionality to one or more 3rd party cloud service providers.  Such service plans are often characterized by a pay-per-use delivery model.

Private Cloud

Another option is to construct your own private cloud, were you use a virtualized IT infrastructure to host services for your own users.  This allows you maintain full control of your own data and the services offered.

A Little Bit of Both

A hybrid cloud solution uses some 3rd party resources, but integrates these with a private cloud for internal applications and services.

The Advantages of Cloud Computing

There are many well-documented advantages to moving some or all of your IT resources to the cloud. Using a cloud-based architecture allows data center managers to:

• rapidly scale resources to meet growing demand
• more effectively support a mobile workforce and enable remote collaboration
• simplify maintenance – less hardware means a smaller support staff
• simplify updates – centralized software management allows IT staff to update applications quickly and easily
• rapidly provision and deploy new technologies
• avoid over provisioning, since computing power is allocated to different tasks in real time

All of these advantages add up to a reduced total cost of ownership for organizations that use the cloud effectively.  Plus, because there’s less hardware to power and cool, cloud computing is often greener than traditional distributed networking.

The Risks of Cloud Computing

Of course, all these advantages don’t come without some risk.

Organizational leadership is primarily concerned with availability.  When software and data storage are outsourced to third parties, there is always a concern that circumstances outside your control will affect access to critical data, either in the form of server latency or downtime.

Front of mind for IT professionals is security.  Hosting applications and data with 3rd party service providers means you have to rely on those providers’ security measures to protect valuable information.  Security is also an issue in a private cloud, since concentrating virtual machines on fewer physical servers magnifies the impact of any single security breach.

Finally, from an infrastructure perspective, variable demands on cloud resources will result in rapid changes in power consumption and the need for heat dispersal.  If your data center infrastructure is not able to respond to these changes, you could face an increased risk of catastrophic system failure.

Making the Call

Before making the decision to move significant IT resources to an external, internal or hybrid cloud, it’s important to optimize your existing facility, including power, precision cooling and network infrastructure.

All too often, data center managers look to cloud computing as a way to expand an overburdened, inefficient data center without additional capital expenditures.  But approaching such a dramatic infrastructure shift in this manner is a recipe for an overcomplicated system that will not perform well or achieve many of the advantages outlined above.

Before making the call, first look for ways to optimize your existing data center.

• Reduce costs
• Improve IT performance
• Improve efficiency
• Refresh hardware
• Right-size their support systems to match their IT systems
• Improve reliability
• Do more with less

As they work hard to meet these sometimes-competing objectives, many data center managers are looking toward co-location providers to enable them to accomplish some of their goals.

A hypothetical example

A government agency has three data centers in its HQ office space.  Looking back, it would have made perfect sense to consolidate all three of these data centers years ago, but turf battles, budgets and politics all got in the way.  Now the agency has three sets of infrastructure to maintain, not to mention three operations teams.  To make matters worse, all three data centers are more than 10 years old, meaning they are far from optimized, from a space, cooling and power perspective.

What are the options?

Consolidate the Data Centers On-Site

Let’s say this hypothetical agency wants to keep their data center assets “in-house”.  The best way to achieve many of the aforementioned goals would be to create an up-to-date data center space within their own property and migrate the contents of their three data centers into it.  This is certainly a significant undertaking, but it can be accomplished with very little downtime and risk with the appropriate planning and outside support.

The downside to this concept is the up-front cost.  Building a data center is not cheap, and in the federal government today, there is a fine line between career suicide and approving capital funds for a new data center.  Before the decision is made to co-locate outside, this Agency would be wise to look at all of the various financing options.  Capital costs can be aggregated and financed through various lease instruments, thereby spreading out that big capital expense into much more manageable monthly costs.  In many cases, these costs will be less than the financial commitment required with an outside co-location provider.

Move all Three Data Centers to a Co-location Data Center

Let’s assume that our Agency doesn’t mind keeping its data center assets off-site, and is willing to move its systems into a co-location facility.  This is happening more and more today, and CIO level leadership is strongly encouraging this option as a way to comply with the Federal Data Center Consolidation Initiative (FDCCI).  Many also assume that this path will be a shortcut to minimizing downtime, maximizing reliability, right-sizing space, reducing staff levels, and lowering costs.

What’s not to love?

Co-location done right.

Moving your data center to a co-location facility is the right decision for many data center managers.  However, to know for sure, one must ask and answer the following questions:

• Can you virtualize your applications in order to reduce the amount of hardware that must be moved and operated in the new environment?  It is not unreasonable to expect to reduce your IT footprint dramatically when consolidating either in-house, or outside.
• Can you move some applications to the cloud now?  Alternatively, can you create your own cloud within the new hosted environment?
• Are your applications and office network “cloud-ready”?  In other words, since your data will be stored elsewhere, will you have the internal and external network in place to send/receive information with minimal latency so the users do not perceive any decline in performance? Are your applications designed to operate across a WAN link?
• Are you prepared to commit to co-location for the long term?  Moving offices is hard.  Moving data centers is extremely hard, and not something you want to do every year or so.
• Are you actually going to save money when you factor in all of the costs and savings?  When you add in the potentially higher telecommunication expenses, are you truly going to save money over the alternative of consolidating in-house?
• Are you still going to need data center space in your current office environment?  While the space allocation can be drastically reduced, it is not unusual to still require a smaller data center footprint in-house for networking gear and possibly some servers.

The big question that government IT leaders should ask themselves is this: when you look at the big picture, are you doing the right thing for the taxpayers by reducing costs, increasing productivity, improving reliability, and reducing your carbon footprint?  Are you considering co-location for all of the right reasons, or is the driving factor the need to placate Mr. VanRoekel, the Federal CIO, and comply with the FDCCI (Federal Data Center Consolidation Initiative)?

Compu Dynamics can provide the Preventive Maintenance solutions for your Mission Critical HVAC system no matter what brand you may own.  We are the only mechanical service organization in the Mid Atlantic Area that is authorized by APC/Schneider Electric as a Certified Service Partner and by Liebert/Emerson as a Liebert USA Contractor.  We also offer services on products by Stulz/ATS, Data Aire, Trane, Carrier, McQuay, York, and many other manufacturers.

Since 1982, Compu Dynamics has been a leading provider of critical power and precision cooling solutions for data centers, network facilities, server rooms, operations centers, points of presence and other communications facilities. Our manufacturer-certified service and installation technicians deliver turnkey, energy efficient solutions for some of the world’s most demanding information technology clients, 7 x 24 x Forever.

Compu Dynamics’ strong ties with the manufacturing community have enabled the company to remain on the cutting edge in providing state-of-the-art customer solutions.  We offer a Complimentary Site Inspection to assess current risks in your IT spaces at absolutely no charge.  We will identify opportunities for improvement and recommendations to ensure your continued operational readiness.

Whether it’s emergency service, preventive maintenance or a complete data center upgrade, Compu Dynamics responds with speed, precision and professionalism to deliver Uptime in No Time™. To find out how our elite team of mechanical and electrical infrastructure specialists can provide your organization with the Compu Dynamics edge, email info@compu-dynamics.com, call 703-796-6070, or visit www.compu-dynamics.com.

The situation at the Fukushima nuclear plant offers a different lesson. As the crisis continues, some media outlets are beginning to examine the plant’s preventive maintenance records. An article published by MSNBC seems to question the operator’s safety record in light of what appears to be missed inspections on 33 pieces of equipment.

As our prayers go out to the people of Japan, let all of us in the data center industry renew our commitment to better prepare for those events we can’t control, and to prevent those which we can. After all, the data centers in our charge are often responsible for providing – either directly or indirectly – essential support for mission-critical public services.

A recent Uptime Institute study indicates that 70% of data center downtime is caused by human error. Another study on behalf of Emerson reveals that 65% of downtime was the result of UPS battery failure. Training, procedures, and a strict commitment to preventive maintenance can greatly reduce unnecessary service outages, and can also result in faster restoration and recovery when the truly unavoidable occurs.

Take a look at your equipment logs, SOPs and MOPs. When was the last time that you replaced your UPS batteries? How about the air filter in your cooling system? Has the recent recession caused delays in upgrades and crucial procedures that could lead to failure if not addressed?

Remember, a chain is only as strong as its weakest link. If you’re unsure which issues are most critical, or would like an unbiased assessment of your maintenance procedures, please give us a call.

Compu Dynamics is a proud supporter of the American Red Cross. We encourage our fellow data center professionals to contribute to the Japan Earthquake and Pacific Tsunami fund at the Red Cross.

1. Evaluate your energy efficiency.

Add up your energy bills, learn where your energy comes from, and how it’s being distributed in your company. Plan a road map for monitoring and reducing usage. Your initial evaluation can provide a baseline for calculating your ROI after implementing changes.

2. Redesign your cooling system.

Enclose your servers in hot aisles so you can concentrate your cooling directly onto the racks. Be sure to reduce airflow loss by sealing off holes in your walls, floors and ceilings, and make sure you have proper air ducts.

3. Reconsider your redundancy.

Many companies invest in redundant cooling and heating systems that use double the amount of power they usually need — for each kilowatt of energy needed, redundant circuits use two. This helps prepare systems for growth and times of peak performance, but the result is a maximum power flow where only a fraction of that power is put to use. This practice is inefficient and inflates your energy bills. Cut back your redundancy wherever possible, to operate based on your current needs rather than projecting your future needs.

4. Use adjustable equipment.

You can also plan for growth and times of peak power needs by implementing scalable and modular systems. Scalable blade servers can concentrate more computing power in a smaller space, requiring less power to cool. Also any scalable system will let you operate at your minimum power requirements and give you the capacity to grow into higher power consumption when you’re ready.

5. Virtualize your storage.

Most servers are underused, because they’re loaded with archaic information or software that’s only needed some of the time. With storage virtualization software or a device, you can make your applications mobile and thereby shutdown unneeded servers. The virtualization device will keep track of your server space and remap applications to different physical locations as necessary to achieve optimal efficiency. Thus, information can be made location independent and redirected across multiple I/O devices from different vendors.

6. Use Energy Star-rated appliances.

Although the EPA hasn’t yet released Energy Star standards for data servers, you can purchase other appliances, such as cooling systems, that do meet current standards. Keep an eye on the industry — new star ratings are due for desktop computers in July, and the EPA may also enforce star ratings for data center appliances sometime in the next year.

7. Donate or recycle legacy servers.

Archaic systems are less efficient than new ones, so you can save money by buying new systems and decommissioning old ones. However, because servers contain huge amount of toxic waste, never throw a machine in the trash. Many nonprofit or educational organizations can make use of old servers. There are also many vendors who handle computer equipment recycling, who can strip the machines of useful parts and properly dispose of the toxic remains.

8. Look outside the datacenter.

Your company can cut costs outside its server rooms, for example by using earth-friendly construction materials such as low-emission carpets and paints and non-toxic insulation; creating skylights, windows and other natural lighting sources; and by landscaping with native and drought-tolerant plants that require less water consumption and support the local ecosystem.

9. Research alternative power sources.

The best way to stay clean is to find alternative energy that works for you. Photovoltaic systems are popular — solar panels on your roof are a simple step and will reduce your dependence on the power grid. Other options include wind energy, heat pumps, and evaporative cooling can lower your cost and significantly reduce your company’s environmental footprint.

10. Get your management involved.

To fully embrace changes in the datacenter, you need to get the buy-in from your upper management. You can get them fired up about being an eco-friendly company by brainstorming earth-friendly ideas that will help everyone in the company, such as using hybrid or electric company vehicles or starting a commuter shuttle for your employees. Once they’re interested, you can easily bring up the green question whenever you discuss the growth of your data center.

A rather remarkable 65% of data center managers survey saw UPS failure as the largest cause of data center downtime.   UPS capacity was seen as the culprit 53% of the time and human error was sited at 51%.

450 data center operation professionals were interviewed.  They reported an  complete outage occurred 2.48 times over a two year period.  Even the big boys, centers with more than 20,000 square feet saw a total failure rate of 1.75 times over two years.

“This transaction is exciting news for Compu Dynamics and Loudoun Electric’s customers”, said Steve Altizer, president of Compu Dynamics. “The high value and trust that Loudoun Electric has built over the years matches our own focused commitment to deliver the most reliable power, cooling, and environmental system design-build solutions for our customers. Now, we will have a full range of in-house electrical capabilities enabling us to provide cost-competitive solutions for emergency back-up systems, UPS power conditioning and electrical distribution up to and including the server cabinet.”

“Our experience and expertise will bring many strategic benefits to Compu Dynamics and to our customers,” said Lee Piazza, owner of Loudoun Electric. “Both Loudoun Electric and Compu Dynamics have earned reputations for excellence and we look forward to joining together in that on-going commitment.”

Today’s data center and facility managers – whether in the public or private sector – face many challenges in keeping mission-critical systems up and running 24×7. Data center consolidation initiatives, capacity scalability, high density applications, preventive maintenance, staffing, green initiatives, remote monitoring and physical security all play vital rolls for today’s data intensive Federal agencies and private organizations. “Our self-perform capabilities, manufacturer certifications and proven design-build expertise incorporating highly energy- efficient systems, give customers peace of mind that our teams can handle their most complex IT needs from a single trusted source. We have solutions for virtually any IT infrastructure problem,” said Altizer. “With Compu Dynamics’ three GSA Schedules, we can satisfy almost any data center requirement.”

For more information on Compu Dynamics world-class data center infrastructure solutions, visit www.compu-dynamics.com or call 703.796.6070.