Building a modern Datacenter

By Scott D. Lowe, James Green and David M. Davis | March 21, 2016

Published with permission of Atlantis Computing

Making Your Data Center Agile: Principles & Strategies
The modern enterprise is a very different beast than the enterprise of days past. In the modern enterprise, IT and the data center play pivotal roles in the operation and success of the business. In fact, the role that IT and the data center plays in helping the business achieve its goals is becoming increasingly critical. IT is not just an expense that impacts the bottom line. For many businesses, IT is a key top-line revenue driver.

However, the old ways of managing these critical resources are no longer sufficient. Rather than build complex conglomerations of components that are carefully crafted into crazy combinations, the emerging enterprise must create environments that are agile and that imbue the business with ferocious flexibility.

There are three key principles around which agile IT revolves:

Thinking big

Starting small

Moving fast

Each of these three principles is discussed below, along with strategies that can help you forge your own path through the modern data center’s forest of options.

Think Big
Small thinkers need not apply. Today’s technology environments demand big thinking from people who have deep understanding of both the business and technology.

Be bold.
In practical terms, here’s what this means: it’s time to look at the whole technology environment and re-evaluate everything.
Does the data center lend itself to emerging constructs, such as pay-as-you-go adoption? Data centers of days past required massive initial investments, which were neither economical, nor particularly practical. Organizations then spent years attempting to recoup these large investments, only to find that they may have never really realized a maximum return.

New thinking may require that you jettison your old ideas around how IT systems are procured, deployed and supported. Your efforts, though, will result in a streamlined IT function that is laser-focused on the needs of the business.

Are you still providing break/fix services for devices like desktop computers and printers? Anyone can do that. Outsource it. You will likely save money and end up getting better service.

It’s easy to say that you should outsource something, but it’s important to understand the reason why you should outsource these kinds of commodity services. That reason is summed up by two words: opportunity cost.

The future of IT revolves around “value added” services, not commodity services that are easily outsourced. For every commodity service that you keep in-house, and to which you dedicate internal staff, there is less time that can be devoted to those value-add projects that can propel the business forward. This time you take away from potential value-add projects is the opportunity cost of the commodity services.

When you’re making the decision to keep a particular service in-house or outsource it, the actual cost of the service is only a part of the decision process. Even in the unlikely scenario that you can perform certain services for less money than an outside party that specializes in those services, you are probably paying a steep opportunity cost to do so.

Although agile IT principles demand that you consider all aspects of the IT organization, since the focus of software defined storage (SDS) and hyperconverged infrastructure (HCI) is on the data center, let’s start there.

Do you have automation and orchestration tools in place that enable a streamlined and more efficient data center? If not, rethink how everything works in your organization. If you have people performing manual, repetitive tasks, look for ways to automate those tasks.

If you have users who would be better served through the implementation of self-service tools, implement them. Just remember, doing so requires that you have the aforementioned automation and orchestration tools in place.

Agile IT also demands that the underlying infrastructure be bereft of complexity in all ways. Traditional infrastructure solutions introduced a great deal of complexity. SDS and HCI solutions bring simplicity to the data center. By bringing ease-of-scale and ease-of-management to what have been very difficult to manage, size, and scale resources, these approaches are perfect infrastructure fits for environments that embrace agile IT principles.

Start Small
With an understanding for what agile IT demands, the next question you may have is, “How do I do it?”

Even the creators of the most ambitious plans need to start somewhere. It’s not generally feasible to simply throw away everything that already exists and replace it with something new. At the very least, new initiatives need to be staged for implementation in such a way that they minimize impact to production.

So, start small. Find something that needs to be fixed and, well, fix it. Perhaps the physical server environment is not well-managed and you have a desire to increase your use of virtualization. At the same time, part of your “big thinking” plan is to move the entire data center to an HCI solution or one that leverages SDS.

Unless you’re a very small company, you can’t do that all at once. So, use the server virtualization initiative as your stepping stone.

Implement your brand new virtual desktop environment on an HCI or SDS solution. Then, during the replacement cycle for your server environment, transition that environment over to the new one. In this way, you’ll maximize your investment in your legacy infrastructure while eventually gaining the benefits of the new one, such as easy scale, fast deployment, and simple administration.

Your projects should be use-case driven and tied to clear business outcomes. Even achieving a simpler, more cost effective data center is a business outcome if it means that the IT department as a whole can now be more responsive to business needs.
Beyond the infrastructure though, starting small may mean finding a place to begin automation. Many organizations, for example, continue to manually provision user accounts. You don’t necessarily have to adopt an expensive and complex identity management solution, but with just a few scripts, you can probably automate a good chunk of the process. The role of the person who used to provision accounts then moves to one of oversight rather than action, freeing that person up to service other higher priority needs.

To be sure, it can be difficult to get started on such initiatives, particularly in IT organizations that are strapped for people resources. However, every automation that is accomplished clears a little of that challenge. The more you can automate, the faster you can automate even more.

Now, with that extra time that you gain in IT, what do you do with it? You improve the business. Meet with the business unit leaders in your company and help them discover and implement new services that increase customer satisfaction or revenues Help them find technology-driven ways to reduce costs. Figuring out how to provide analytics that helps a company market better to their customers. For example,if you’re a bank, helping to implement the ability to scan checks to deposit them via smartphones is a great example of how technology can drive customer satisfaction. This service changed the retail banking industry and it’s primarily a technology solution.

Move Fast
If you’re asking yourself when you should get started, that’s an easy answer: Now! Don’t wait.

Here are six steps to accomplish your goals quickly:

1. Talk to the Business.
Interview business unit leaders and senior management and identify business objectives and where technology support gaps may exist or where there may be opportunity to use technology to improve a process or introduce a new service.

2. Assess the Technology Environment.
At this point, you know what the business needs. The next question is this: can the technology environment support it?

The technology environment (the hardware and software) itself is absolutely critical to the business. However, over time, as is the case with many things, such environments begin to take on lives of their own. What once was state-of-the-art is now a complex morass that has been extended to meet unanticipated needs. Of course, this is not always the case, but it’s certainly not unusual, either.

The more that the environment has been customized and extended with a lot of different point solutions, the more complex it is to manage and the slower that IT can react in modifying it to meet new business demands.

Your job is to figure out which technology solutions need to be maintained in order to meet business needs and then find ways to simplify what remains so that it is manageable and cost effective. This need absolutely includes the data center, too.

Look at the data center environment both holistically and granularly. With all of the information you gather, create a list of every product and service and identify, where appropriate, the business need served by that product or service. For example, your marketing department might have its own application environment that supports their customer relationship management (CRM) system.

Your job at this point is not to eliminate services and products, but simply to identify.

3. Create a Support Inventory.
Analyze every aspect of the IT organization and list every process that is currently supported and determine where there are support gaps based on what you learn in the previous step.

For example, is that CRM environment being maintained in a way that truly meets the needs of the business? If not, why not? Is it because the staff has skill deficiencies? Is it because there are too few staff to support the existing operational environment? Is it because the existing staff do not have a focus on the business?

This is your opportunity to determine where there may be deficiencies that would prevent IT from fully executing on the strategic priorities that were identified during your discussions outlined in the previous steps. In this step, focus on people and process.

4. Identify Core Services and Gaps.
At this point, you should have a good understanding for how the various products and services present in the organization actually meet business needs. You should also have a good understanding for where gaps exist in the environment.

You should be able to answer the following questions:

Which products and services are core to the mission and vision of the business?
What gaps exist in the current product portfolio that make it difficult for IT to meet business goals?
What gaps exist in the IT support structure that make it difficult for the business to meet its goals?

The core services list may not mean that you need to maintain the status quo. It simply means that the particular service is required to be provided in some way. It’s here that you begin taking concrete action to correct deficiencies and help propel IT forward.

5. Decide What to Outsource.
Many IT departments continue to believe that all technology services must be provided by the in-house IT department. That is simply not the case anymore.

People have been really wary of the word outsource because it can imply that people will lose their jobs. But you have to look beyond all of this and figure out what is best for the business. Outsourcing is not just about saving money; it’s also about creating opportunity for the business. Every time you outsource a function, the internal people that were handling that function can redirect their efforts to more value-add projects.

As you review the list of services being supported by IT and the list of support gaps that you have identified, you need to look for the best way to address the core services. Sometimes, this will mean outsourcing a core service. For example, if Microsoft Exchange is a core service, does it make more sense to keep it in- house or should you move it to Office 365?

Other times, it may mean outsourcing a non-core service. For example, you might choose to outsource your SharePoint portal and redirect those staffing resources to managing your Exchange infrastructure.

With your lists in hand, figure out what you can “make someone else’s problem” via outsourcing and then pick up the phone and start calling potential vendors.

6. Decide What to Improve.
With what’s left, you must now decide where you want to improve. The hard part can be figuring out where to begin. With that in mind, here’s a tip

Look at the items on the list and, for each one, assign two metrics (a value of 1 to 5 for each):

The difficulty level for addressing that need. For the difficulty metric, a 1 indicates that it’s low difficulty and a 5 means that it’s a high level of difficulty.

cost metric, a 1 means low payoff value while 5 means high payoff value.

Now, immediately attack the items that have a 1 or 2 difficulty level and a 4 or 5 payoff level. Get them done and done fast.
Obviously, you will need to coordinate with business leaders to really prioritize your efforts so that everything you do aligns with what the business needs.

Conclusion
In this chapter, you learned that time is not on your side. But by leveraging the tools and services available to you, it’s possible to move beyond just “keeping the lights on.” Following are a few key terms and concepts to take away.

Agile IT is a set of principles that require that people think big, start small, and move fast.