Tuesday, October 31, 2006

The Pros and Cons of Data Center Colocation

Data center colocation plays a strategic role in modern IT operations. For many companies, it presents an attractive alternative to building and maintaining a sizeable data center.

Housing servers and network equipment off-site does offer companies a number of benefits, but the picture isn’t as simple as black and white. Before you commit to colocation, here are a few things to consider:

Data center colocation allows you to…
- Eliminate the up-front costs associated with building your own data center,
- Deploy IT infrastructure based on your company’s current network requirements with less impact to your budget, and
- Relocate the company without subjecting your network to downtime or other interruptions.

Data center colocation becomes a less attractive option when…
- The overall space and load requirements are substantial,
- Your data center requires intensive management, and
- Physical security is a major concern for your company.

Data center colocation providers can charge based on a number of requirements, including total physical space, amount of bandwidth, amount of UPS protection, quantity and type of power distribution, and the degree of physical security needed. That’s in addition to the cost of other managed services that take the place of your in-house staffing and tools. It’s easy to see how the costs can add up if your data center is high-maintenance in one or more of these areas.

Colocation is not a one-size-fits-all solution for data centers. Much like choosing between buying your own house versus renting an apartment, the decision to maintain your own data center or hire a colocation provider should be weighed based on your company’s present and future needs. Companies often benefit from colocation in the short term, but the option to house servers and network equipment off-site needs to be evaluated based on your big picture.

Wednesday, October 11, 2006

Datacenter Solution: Server Virtualization

You’ve probably heard the phrase, “Don’t put all your eggs in one basket.” Does this folksy wisdom hold true when it comes to server virtualization? Virtualization is an increasingly popular datacenter solution that divides the resources of your physical server into multiple isolated virtual environments. Instead of maintaining one application per physical server, virtualization enables dominant workloads to run simultaneously on the same server.

Although conventional wisdom tells us that putting all your eggs in one basket is fool-hardy idea, carrying a bunch of baskets around doesn’t seem like a smart solution either. Should you consolidate many physical servers into fewer servers hosting virtual machines or would you be better off maintaining a 1:1 ratio of servers to applications? Deciding which solution is the best for your datacenter depends on a number of factors. To help you weigh your options, here are some pros and cons of server virtualization:

Reasons to Keep Your Eggs in One Basket:
The Benefits of Server Virtualization

Carrying one basket is better than carrying several mostly-empty baskets.
- Virtualization makes better use of datacenter resources by utilizing more of the available capacity on each server, thereby increasing the operational efficiency of your datacenter.

You don’t want to use more than one arm to carry your eggs.
- By reducing the number of servers that are required to run your applications, you improve the manageability of your facility. This means that fewer administrators are needed, allowing you to reduce costs.

You don’t have space for all those other baskets.
- Virtualization enables datacenters to reclaim space in the server room and allows companies to better handle future growth without high capital cost support infrastructure changes. Instead of adding a new physical server for each new operating system, the number of servers can be reduced while the number of operating systems stays the same or increases.

Reasons NOT to Keep Your Eggs in One Basket:
Problems with Server Virtualization

If you drop your basket, all of the eggs could break.
- With multiple applications hosted on each server, a catastrophic failure in a single server could cripple your operations. The cost of establishing an adequate contingency plan should be factored into your virtualization strategy from the beginning in order to make this datacenter solution effective.

An egg-filled basket is heavy – you’d need to invest in a stronger basket.
- If all your eggs are in just a few baskets, you want to make sure those baskets are sturdy. Many organizations use inexpensive servers without high-availability (HA) components for lightly-used applications. Once virtualized, a server must be HA in order to support the mission critical load. Thus virtualization often increases the overall cost of each server while reducing the quantity of servers.

You might need to hire someone to help carry that big basket of eggs.
- Before introducing virtualization technology, you’ll need to send your IT staff for training in virtualization deployment and management, or hire an outside firm to provide virtualization support. Also, while virtualization may reduce your overall hardware costs, the datacenter solution increases the complexity of your operations. This introduces new support costs since you have to maintain the virtualization software in addition to your existing operating system and applications.


Egg-basket metaphors aside, the decision to enhance your datacenter using a virtual machine model is a matter of weighing the benefits against the expenses of virtualization. Server virtualization is not a one-size-fits-all datacenter solution, but it’s an approach to consolidation that many businesses may find beneficial.

Friday, September 29, 2006

Tips for Handling Your Data Center Relocation

Nobody ever said data center relocation was easy. Although moving mission critical computer and communications equipment from one site to another is no simple task, it needn’t be an overwhelming one. With careful planning, any company can organize its people, manage the communications, implement a sophisticated design for the new site and move its data center equipment with confidence and ease.

Months of planning typically go into planning the successful relocation of a data center. By placing an emphasis on pre-design and planning, data center managers and relocation consultants are able to achieve the optimal solution to meet the demands of even the most complex relocation.

If your company is contemplating a data center relocation project, the following tips will help ensure that your IT move goes off without a hitch:

- Invest in pre-planning.
Identify your operational objectives and use them as a reference point in planning. Coordinate with vendors who will be working on the project and include their representatives in your planning sessions. Consider hiring an experienced data center relocation consultant as your guide.

- Develop strategies that minimize risk and maximize ROI.
Explore the backup and recovery options for your mission critical systems. Take current and future needs into consideration while designing the data center at your new site.

- Take stock of data center equipment.
Data center relocations offer the perfect opportunity to replace outdated or faulty equipment with state-of-the-art hardware. Don’t spend valuable time and money moving equipment that won’t last.

- Test the network.
Run tests of your new data center on virtual models before construction and relocation begin to prevent unfortunate surprises. After the move, run additional network operations and applications tests to protect your IT systems.

- Update security procedures and technology-use policies.
Address any new security issues raised by the new data center site. Establish guidelines that safeguard your facility from credible threats and existing vulnerabilities.

- Establish an appropriate relocation plan.
Integrate redundancies into your IT network. Alert business users to potential service disruptions. Review the relocation strategy with vendors, consultants, and other special participants before moving ahead.

By keeping these points in mind, your company can minimize relocation-related downtime while streamlining the overall data center moving procedure.

Thursday, September 14, 2006

DC-power Sparks Interest among Data Center Designers

Rising data center power and cooling costs have renewed industry interest in DC-power, but is DC really a cure-all or just more snake oil?

Up until this point, DC-power hasn’t been very attractive for a number of reasons. The large amount of copper wiring required for DC-power distribution is expensive and takes up valuable space in the data center. DC-engineered facilities are also more complicated to install and maintain, creating headaches that most data center professionals don’t want or need.

However, tech industry leaders, such as Sun Microsystems and Intel, are paying attention to DC-power. They know data centers are desperate to cut power and cooling costs and are banking that many larger data centers can afford the upfront investment in DC-engineered facilities.

This summer, more than 20 tech companies teamed up with the Department of Energy’s Lawrence Berkeley National Laboratory to run a side-by-side demonstration of a conventional AC-powered rack of servers next to a DC-powered rack. The hope was that DC might prove to be the cheaper, cooler-running, and more reliable data center power option.

AC vs. DC

Could DC-power have finally come of age? According to a whitepaper by American Power Conversion Corp., one of our trusted data center vendors, AC-power and DC-power should be compared based on efficiency, cost, compatibility and reliability. Let’s use this as our guide for comparison.

Efficiency:
DC-power is often touted as being more efficient than AC-power because it requires fewer power conversions that can result in power losses. On the surface this makes sense, but it doesn’t take into consideration that DC cabling must be greatly oversized to deliver the same power. You end up trading increased infrastructure costs and a loss of space for a relatively minor gain in data center efficiency.

Cost:
There isn’t a real cost advantage to making the switch to DC, even though DC-power might lower your overall energy costs. As I mentioned earlier, DC-power requires more extensive infrastructure and engineering. There’s a higher upfront cost which offsets most long-term savings.

Compatibility:
This one is a toss-up as well. Compatibility with AC- or DC-power supplies is determined largely by the type of equipment your facility uses. Keep in mind, though, that it is difficult to obtain DC versions of many pieces of equipment. If your data center uses a mix of AC and DC products, it’s probably a wise decision to go with AC-power.

Reliability:
For the most part, reliability is controlled by battery system rather than the type of power used. When compared with an equivalent DC-powered system, AC-power is just as reliable for data center use.

PTS’ Perspective

The use of DC-power in the data center isn’t common – and with good reason. There is no clear benefit to switching from AC- to DC-power, but there are quite a few drawbacks. Rising energy costs are enough to make companies reevaluate the concept, but AC will remain the dominant choice for data center power in the foreseeable future.

Thursday, August 31, 2006

Data Centers vs. Computer Rooms: What’s the Difference?

The differences between data center and computer room design don’t amount to a hill of beans for most people. The terms are often used interchangeably, but using them correctly makes a big difference if you’re trying to communicate with a data center design firm or an IT expert. If you want to sound like a pro, it’s important to know what sets data centers and computer rooms apart.

Data centers are designed to provide a secure, power protected, environmentally controlled space used for housing server, network and computer equipment. As the operating theatre for an enterprise’s network service delivery, a data center site may utilize the entire site and building shell.

The design of computer rooms is more limited in scope. A computer room is merely a functional space within a data center. It serves as a secure environment for the equipment and cabling directly related to the critical load. In other words, a computer room’s basic design is that of a collapsed data center where the entrance room is contained within the computer room space.

The easiest way to tell the design of a data center from that of a computer room is by looking at how the space’s functional pieces are put together. A data center is a larger space composed of smaller spaces, such as a computer room, network operations center, staging area and conference rooms.

In either case, data center design and computer room design are both accomplished by identifying the key design criteria for the two main areas of the project focus – the technology infrastructure and services (IT) and the support infrastructure and services (the facility). The key design criteria are:
- Business Objects (Scope)
- Availability Requirement
- Power and Cooling Density

While site selection is also a criterion for data center projects, a computer room design project can be as involved as a bigger base-building project or as simple as an upgrade of an existing computer room.

Understanding the differences between data centers and computer rooms is the first step on the road to delivering a successful data center or computer room project. The more you know about the elements of a data center, the easier it will be for you to get your design ideas across to others. If you’d like to learn more about this topic or others, we invite you to visit our White Paper archive at http://www.pts-media.com (registration required) or contact us.

Friday, August 25, 2006

Dark Data Centers: Dream or Reality?

I was just contacted by Processor.com for my thoughts on this topic and thought it might be useful to share some information about it with you all here.

If data center operators had only one wish, it would be this: build me a dark data center.

Many of the daily problems that affect data centers have less to do with the design of a facility and more to do with variables introduced by human involvement. For most data centers, not only does the IT staff have access to the facility. Facility staff, other employees, outside consultants, contractors, and mechanics may enter the data center for a whole host of reasons. As human traffic increases within the data center, so do the risks, amount of clutter, and the number of potential technical problems.

Despite expert design and planning, people do not always follow preset procedures and may meddle with equipment that they are not qualified to use. This is a nightmare for IT professionals. The mistakes are difficult to trace and are a drain on a business’ money and the time of its IT staff.

The ideal solution is to design a dark data center, a remotely monitored IT environment, in which computer systems analyze and correct problems with minimal human involvement. To achieve a completely dark data center, your IT infrastructure, support infrastructure, and software systems need to be autonomous. The majority of companies are no where near this point and most data centers will never be able to run without any human interaction, but technology is quickly taking us closer to this design goal.

Not-So-Dark Data Center Design

Cutting the human element entirely out of the picture may be out of our current reach, but you can reduce foot traffic and the number of unmanaged changes within your data center. “Dim” data center designs are a realistic goal for most companies.

The dim data center approach focuses more on preventative maintenance than on reactive problem solving. The most effective dim data center designs are secure, can independently troubleshoot most problems, can be managed remotely, and implement processes and procedures to control the who, what, where, and when’s of the events taking place within the space.

Dim data centers remain a sought-after solution for IT professionals and users, and an attainable design goal for most companies. Although the dark data center is still a dream, the dim data center is a happy reality.

Tuesday, August 22, 2006

Data Centers Go to Washington

Data center power and cooling issues are creating quite a buzz on Capitol Hill.

In recent years, the power and cooling costs of the average data center have gone through the roof. Data centers rack up more than $3-billion in energy costs each year. That number is expected to rise dramatically within the next decade as more data centers are built. Adding to the energy drain are factors such as inefficient cooling systems, more powerful servers, and rising energy prices.

Part of the problem is that there isn’t enough energy to go around. This is a bigger issue during the summer months as the nation’s reserve electric capacity is declining. Some utility companies have asked business customers to cut power usage during peak times, even if that means switching from the power grid to a generator.

Congress’ aim is to pass legislation that will help to promote the use of energy efficient technology. This past July, the House passed a resolution that instructs the federal Environmental Protection Agency (EPA) to study the issues surrounding data center power consumption, assess what the industry is doing to develop more energy-efficient technology, and seek incentives that would encourage companies to make the switch.

Now it’s the Senate’s turn to address the issue. In the hope of finding ways to cut back on the amount of power consumed by corporate and federal data centers, the Senate introduced a bill that is nearly identical to the one passed by the House just a few weeks earlier. If the data center power efficiency bill passes, it will go before President Bush for his signature.

Industry Reaction

So far, industry reaction has been more positive than negative. Congressional legislation is seen as an important step in raising national awareness of data center power issues, although some technology professionals are worried that this could lead to unnecessary regulation.

Rather than having the federal government set down rules for how data centers are designed, industry insiders hope the data center power efficiency bill will result in something similar to the Energy Star rating seen on computers and appliances. Such a rating would encourage manufacturers to improve the efficiency of technology without stifling industry growth.

PTS’ Perspective

Whether or not the legislation has a profound impact on the data center industry, there are steps that your organization can take to improve efficiency and save on data center power costs:

- Choose the most energy efficient data processing equipment. According to AMD’s Tony DiColli, AMD’s Opteron processor can consume 27% to 80% less power than its Intel Xenon counterpart.

- Use scalable modular support infrastructure that allows the data center power and cooling infrastructure to grow with the load, rather than over-sizing your data center to compensate for future growth.

- Improve the efficiency of your cooling systems. The cooling load should include both the IT load and the room heat load components including skin loads, lighting, people, outside air sources, and heat dissipated due to inefficiency of power and cooling components.

- Reduce the non-critical load losses of power and cooling components in the data center. These are losses that are independent of the load, such as control logic losses.

For an expert analysis of your data center’s power efficiency, consult a data center design firm. By looking at your IT environment as a whole, data center design professionals can weigh the complex interactions of your facility's elements and provide recommendations for improvement.

If you’d like to learn more about data center power efficiency, please read “Electrical Efficiency Modeling of Data Centers” by Neil Rasmussen. A complimentary copy of the White Paper can be downloaded at PTSDCS.com.

Thursday, August 17, 2006

Data Center Solutions to Beat the Summer Heat

This summer, data center managers are sweating – and it’s not just because of the heat. As temperatures hit record highs this year, the increased demands placed on US energy grids have lead to brownouts and occasional blackouts. Between the soaring temperatures and the power outages, data center cooling systems and backup generators are getting a real work-out.

The situation isn’t likely to improve in the summers to come. As our national energy usage increases, power reliability decreases during peak times. The demand for energy is growing at a much faster rate than our power generation capacity can handle.

Unfortunately, if you can’t take the heat, the solution isn’t “Get out of the data center.” Data centers that don’t want to get burned by future heat waves are investing in some preventative measures. Many businesses are installing extra cooling systems and on-site generators in the hope that these data center solutions will prevent costly downtime.

Heat removal is essential to the proper functioning of data centers, yet poor design and maintenance choices prevent many air conditioning systems from operating at peak efficiency. The availability and reliability of your network services hinge on the continued operation of your precision cooling solutions. If you’re looking for help minimizing the frequency and severity of unexpected downtime, try the following data center solutions:

- Provide redundancy throughout the entire cooling infrastructure by maintaining at least one additional computer room air conditioner (CRAC), pumps, and heat rejection equipment for each cooling zone. This is referred to as N+1 redundant.

- If you have a cooling system that employs on-site thermal storage such as a chilled water system, consider providing the air handlers, inside the computer room, with uninterruptible power supply (UPS) power to provide uninterrupted power & cooling to the site.

- To sidestep power outages altogether, size the on-site emergency power generators to handle your system’s cooling as well as power needs.

- Perform regular checks on your computer room air conditioner (CRAC) and heat rejection equipment including inspecting all filters and operating parameters.

For an expert assessment of your data center’s cooling system, consult a data center design firm. By evaluating your present and future loads, capacity and redundant capacity plans can be created and utilized to keep you cool when the heat turns up.

Monday, May 08, 2006

Raised Floors - Still Relevant?

With the widespread use of serves installed in racks instead of legacy mainframe configurations, the use of ethernet cable in cable trays located above server racks instead of legacy wiring types beetween mainframe boxes, and finally the use of in-line cooling equipment and/or overhead cooling with ducted returns for cold aisle/hot aisle configurations, is the installation of an access floor a wise use of money when developing a computer room?

Tuesday, April 18, 2006

Data Center Planning is Critical to a Successful Project

Network downtime means lost revenue – and lost jobs
Imagine a scenario where you are coming under fire because your data center project is over budget and months late. Even worse, your management team pushed you to accept the architect’s idea of your computer room that today won’t even deliver the most basic needs - enough space, adequate security & fire protection, ample power, and sufficient cooling.
Now picture a different scenario. But this time, you have earned the respect and appreciation of your company by having properly planned, designed, and delivered an always available data center project. Within scope. On time. And within budget.

What makes the difference in these two scenarios? With PTS, when your management team came to you expecting you to provide the new computer room’s power and cooling requirements, you had a data center design expert you turned to - an expert who had your best interests in mind. An expert who understood the obstacles in providing constant network availability, translated how future IT services impacted the critical load, and conceptualized how the resulting supporting infrastructure impacted the budget.

A Good Plan Pays for Itself
We know it’s your job to deliver on management’s expectations of always available network services. Further, we know that you can’t be accomplished that without a room and supporting infrastructure equal to the task. PTS’s data center planning, and pre-design, consulting services ensure your company gets the always available data center it requires – from the start.
  • Clear project cost guidance as it relates to the overall availability expectation
  • Accurate existing equipment & conceptual IT services critical load profiling
  • Expert understanding of data center design standard, and not-so-standard, practices.

A good plan is the key to delivering a successful project. And a good project manager prevents as many things from going wrong as possible by using that plan.

At PTS, we’re in the planning business along with you.

Cost Versus Availability
Everyone wants a 7x24x365 environment. However, the impact of delivering that kind of performance is far-reaching. Availability is more than the reliability of components used and the redundancy by which they are configured. Systems must be concurrently maintainable as well. This means that as absolutely necessary periodic maintenance is performed, systems must be able to be completely brought ‘off-line’ without impacting the load.
Not surprisingly, the greater the availability expectation, the greater the cost.

PTS helps you:

  • Determine the availability that is appropriate.
  • Estimate potential costs at that risk level.

Accurate Critical Load Profiling
Most architects and consulting engineers rely on their IT staff to provide space as well as equipment load requirements to determine the required facility infrastructure. PTS translates IT equipment details directly into power, cooling, and space requirements for the facility. We have found that equipment nameplate or manufacturer specifications are insufficient in many cases and we have compiled our own techniques for such purposes.

Additionally, PTS has the project experience necessary to provide accurate load profiles and their facility infrastructure impact from conceptual IT initiatives and services without detailed equipment lists. This includes the impact due to technology refresh initiatives such as blade server deployment, equipment compaction, and more.

Standard and Not-so-Standard Practices
PTS has developed a detailed set of design standards by which your data center availability can be accurately predicted. We understand the obstacles facing uninterrupted data center operations. And we know how to design and implement the solutions to mitigate them. (Many architects and consulting engineers are not fully aware of the latest high density cooling trends and techniques, and the impact they have on facility infrastructure.)

Save Money, Save Time – Save Your Data Center
PTS is uniquely qualified to help businesses assess their data center and support infrastructure risks, provide recommendations for improvement and offer accurate project cost estimating and guidance throughout implementation.

Please contact us and tell us about your project. Give us the opportunity and we’ll provide you with the justification you need to convince your management to take the next step.

For More Information...
Please contact us for more information regarding PTS Data Center Solutions at
1-(866)-PTSDCS1 / 1-(866)-787-3271 / info@ptsdcs.com.

Or visit us online at, www.PTSdcs.com.

Experts for Your Always Available Data Center