Weapon 3: IaaS Can Be Implemented In-House
Infrastructure as a Service is both a structural concept and a mindset. As such it can be implemented internally; it doesn't have to come from an outside service provider. If you implement IaaS internally, it means the IT department can charge-back its services proportionally to the parts of the company that are the heaviest users. Suppose you're in manufacturing and the production side never has any sizeable needs in terms of IT, but marketing, HR, and accounting are crushing you. When budget cycle comes around, the accounting people, who are draining half your resource, won't allow you any increase. How do you align with that? With IaaS, you can charge them for IT as a service, so if they want more, they have to pay more.

Alternatively, another new solution enabled by IaaS is to put your IT budget inside other departments' budgets. You give them dollars to spend versus IT having to carry and justify those expenses throughout the year. So if marketing decides in the eleventh month that it needs to double its storage, you don't have to come up with a solution for which you don't have the money. Marketing now has to pay for it out of its budget.

Weapon 4: More Opportunities Captured with Just-In-Time Resources
How often do businesses pass up opportunities because they don't know the cost or aren't in a position to scale? Earlier I alluded to the difficulty of planning IT expenditures on an annual basis. The problem is you're planning blind in many cases.

From a business growth standpoint, you have to be ready to expand, but don't want to spend big bucks on IT resources until you absolutely have to. Let's say you plan to add two more Fortune 1000 customers next year. Why would you spend the capital or even set it aside on January 1, not knowing whether those accounts are going to close in March or October? Even if you put the money in the budget and great things happen and you close the new business, you still need a rapid way to provision that environment. IaaS can help you ramp up - just outsource the infrastructure until you've built your own capability. You can have entire environments up and running in days, sometimes hours, instead of weeks.

Now factor in serendipity. Instead of adding two large accounts, you add five. Oops! Three accounts to go and your budget is gone. No one can tell you when or if the business will land a third, fourth, or fifth large customer. To handle it, apply the same model. After you land that third big customer, send it to an Infrastructure as a Service provider. Once you get budget approval bring it in-house, pull it back when you're ready. It's that easy.

Weapon 5: Change Is Good When IT Is More Flexible
In business, opportunity implies change, and change can always be a challenge. IaaS enables rapid change because it lets companies add or remove infrastructure and services on-demand. While rapid change can impact stability, with IaaS, you can add horsepower to the 60%-80% of your IT environment that's already stable, while gaining more control over the 20%-40% that's in chaos.

One of the misconceptions about IaaS is that when you decide to use this kind of outsourcing, it's a permanent decision. Imagine if you had to grow your infrastructure 10%-20% in 30 days. If you decided to use a service provider to help, you're not permanently stuck in that mode nor have you set a precedent for the future. One of the ideas behind Infrastructure as a Service is that not only can you scale up quickly, you can also scale down or scale out.

Infrastructure as a Service is generally delivered in addition to a utility computing platform. As long as you have a platform like VMware for virtualization, you look identical to your infrastructure provider. So, if you wanted to push 30 machines or 50 machines out of your data center for 90 days, you could easily bring them back because you're both running the same virtual platform.

Weapon 6: Data Center Automation Kicks IT into High Gear
System administrators in today's data center are typically familiar with only 10-20 specific hosts or devices. These are the units that fall under their area of responsibility and are often their area of expertise. With new technologies such as virtualization and high-density computing, the quantity of actual devices in a data center grows rapidly.

Administrators and engineers have to learn how to manage large groups of like devices as fleets. The days of logging into a single box to run patches, tweak the registry, or change permissions are gone. There simply isn't enough time. Data center automation tools like Opsware (soon to be acquired by HP) and BladeLogic are a large part of the IaaS model because they enable a single administrator to manage potentially hundreds of devices.

These tools provide templates and policies for configuration, patch management, and security compliance. You can configure a single template based on best practices or corporate policy and apply it to several hundred machines. A delta report will show you all of the devices that need attention. Built-in automated remediation lets you select all of the devices and apply a single change or group of changes at once. The tools can group devices (if you have different policies for development versus production, for example) as well as exceptions for patches or changes that haven't been tested yet.

Weapon 7: Achieving Regulatory Compliance Is Easier
If ever two words in the English language were designed to strike fear in an IT director's heart, regulatory compliance are those words. Healthcare has HIPAA, publicly traded companies have SOX and COBIT, the e-commerce world has PCI. Once again, IaaS comes to the rescue. Using the underlying features of IaaS, compliance becomes easier. Pre-compliant virtual machines can be kept in a library, giving you a head start when a new application environment has to be deployed. Instead of installing the server from scratch, you can deploy a copy of a pre-configured (and even pre-compliant) virtual machine. Many organizations maintain a stockpile of pre-built virtual machines in a library for this purpose. It also dramatically improves your provisioning time. After the servers are online, using data center automation templates, you can keep the machines in compliance and even monitor their compliance and patch-level status in a dashboard.

You can even charge more for the cost of a compliant environment as opposed to a standard one to offset the additional management and data center automation software licenses. That way the business can decide and use compliance appropriately. Given the inherent flexibility of an IaaS architecture, it can even change its mind and turn compliance on later.

Weapon 8: Business Continuity Is Assured in the Event of a Disaster
From an earthquake to a virus to a malfunctioning sprinkler system, there are hundreds of ways to lose data and only a few really good ways to recover it. So far, IaaS is the best way I've seen. Data can be backed up automatically in real-time to a strategic network of data centers that serve as mirrored storage and backup sites. Thanks to virtualization you can store multiple backup servers on a single physical server at a backup site and greatly reduce your hardware and operating costs. IaaS providers generally sell these as backup "targets" if you don't want to make a capital spend and put your own equipment in a co-location facility.

If a recovery is declared, you'd initially come up on the consolidated backup hardware, but as performance requirements dictated, your IaaS provider can supply additional hardware on-demand until you're back up to full production status. Because virtual machines are bootable, instead of performing a bare metal restore or reinstall, you simply boot up the VMs, which significantly reduces your Recovery Time Objective (RTO). The virtual machines also contain all of that precious custom configuration information that's so often lost or under-documented.

The use of a virtual approach also reduces the issues of hardware compatibility. As long as the virtual machines will run on VMware and as long as VMware is installed on the recovery hardware, they'll boot. Not only does this help recovery, but it makes testing a lot easier. Most, if not all, of this activity can be done remotely since no tapes are involved. This again saves time because your administrators can avoid the airport and get right down to work.

If you consider the growth of the computer sector, the centrality of data and software to business, and the strain on capital resources that imperfect IT planning can cause, you can see why Infrastructure as a Service is perhaps the ultimate weapon for your success..

• Services Delivery Goes Pervasive