Yes, users should consider migrating to a utility model. But don't tune out yet. And yes, the idea has been floating around the data centre for the past three years, and yes, many people remain sceptical.

But what if the next time you start a project associated with your storage strategy -- especially upgrading your storage network -- you consider employing a utility model as part of the planning process?

In many ways, the process required to build a storage utility isn't that different from what users are already doing in order to gain support for IT budget expenditures. The differences come down to how new storage projects are funded, how administrators manage their operations and how dependent business operations have become on storage and other data center operations.

Much of the confusion over the utility model is due to the feeling that it requires outsourcing some part of operations to a vendor that, in turn, charges back based on capacity or processing usage. While this is one approach, it's not the only one enterprises are taking. In fact, more enterprises will deploy their own internal private utility models that accomplish many of the same goals as outsourcing -- and without giving up the data center keys.

Defining the utility model
So what is a utility model? It's a number of things. It's a deployment model that changes the way storage and other data centre services are managed. It's a new way of accounting for IT infrastructure usage. But most of all, it really is a new philosophy meant to answer the increasingly asked question: How is IT relevant to business?

Traditionally, the IT budget process has included a lot of lobbying by IT administrators. They make their cases for spending a fixed sum for a storage network, or allocating funds for a new storage system or better management software. The pitch also includes asking for money to maintain everything. That's the process.

But imagine a model in which storage and storage networking purchasing decisions made by storage administrators become directly tied to business requirements. It could happen if an innovative IT person decided it was time for storage purchases and operations to be directly tied to business operations. It could also happen if IT managers improved existing operations, such as backup, so they could be run as internal, usage-based charge-back services.

In a lot of cases, people don't even know that some of their newer management and financial practices are, in essence, treating their storage operations as though they were utilities. A number of data centre executives at a financial services forum admitted that, while they were interested in running more of their operations as computing and storage utilities, they didn't have time to put full strategies in place.

Interestingly, a number of these execs were getting direct contributions from business units for their annual IT budgets, and others were already charging business units for usage, capacity and processing. Both of these practices are hallmarks of the utility model that have already become common at many Global 2000 companies.

There are a number of misconceptions about the storage utility model. They include:

  • The enabling technologies are not yet available. While you could argue that some of the management tools that meter usage should be a lot stronger, most currently available technology is good enough to form the foundation of a utility strategy. Things like storage resource management, storage virtualisation and more advanced service managers/management consoles are good starting points. There will be some significant enhancements to technology in the next several years, but there's no reason not to start with a basic model today.
  • It's impossible to build the business case for a utility approach. Yes, it will take work to create the initial business case for developing a utility model, but the argument is greater against not doing it. This is because the payback from new investments, as well as maintenance and labor costs, could have less of an impact on the storage budget if these costs were directly linked to offering a chargeback service. The chargeback approach is also beneficial because it puts pressure on business units to be more judicious in their use of IT resources.
  • It's only available if you want to outsource part or all of your data centres. Actually, outsourcing is only one of the avenues that customers can travel to develop a utility model or philosophy. I believe it's not the best way to get your first taste of what a utility approach can offer -- unless you're confident it will be very cost-effective. In fact, building your own pilot programme gives you a sense of what it will take to gain support from business unit managers, as well as what you should be charging these business units as part of this service model.
  • Business unit executives will never support it. On the contrary, if it makes IT budgets more efficient and reduces business costs over time, business executives should be on board with utility models. Why? Because utility storage and computing models actually change IT operations from being cost centres to fitting in with a business unit philosophy -- complete with profit and liability accountability.
  • So, where do you start?

    There are a number of things users should consider when planning a utility storage strategy. First, do an inventory of your storage operations. What are the costly, business-critical services you currently provide to business units?

    Second, figure out how you would charge the business units for this service.

    Third, make a plan. Key things that need to be considered include: How will you monitor usage? What's the rollout strategy? How do you educate business unit managers about their consumption, and how can they reduce expenses?

    Finally, prepare to change. You won't be able to turn your entire storage budget into a cost-sharing model without a lot of education and hard work. It will take time, so create easy pilots that start to get business types thinking about what IT (and specifically storage) costs really mean to them.

    Following are projects you should consider as possible storage utility model pilots:

  • Update your storage network: Adding a new storage network or augmenting an existing one is one of the most common storage administrator activities. Analyze your current storage network usage. Could new network expenditures, maintenance and ongoing labor costs be absorbed by a business unit that is paying you for services?
  • Backup/recovery upgrades: One of the more common ways to start with a storage utility model is by measuring business unit activity in terms of capacity used. Consider how you could share the cost of upgrading to a new backup platform (both hardware and software) at the business unit level. I have spoken to dozens of IT administrators who are already doing this.
  • Data retention strategies: It makes sense for the business unit that will benefit from this strategy to be responsible for underwriting it. Charging business units for retained data at a capacity level only makes sense.