The Business Case for Storage Networks [Electronic resources] نسخه متنی

Total Cost of Ownership

TCO analysisalthough it is painstaking and time-consumingis the only way to provide the level of detail and coverage required to help business stakeholders understand the impact of using technology. TCO analysis counters the reigning perception that resources are, because of economies of scale and purchasing power, free. Looking at the TCO of a solution is the only way to understand the impact of using or implementing a technology solution. Ultimately, having an accurate TCO number for storage is the only way to determine the ROI from increased utilization and is a fundamental requirement for creating the tiered storage model. Having an up-to-date storage TCO requires significant effort and ongoing analysis, but it is well worth the effort.

A functional TCO for storage reflects all aspects of the costs of implementing a storage solution including all hard and soft costs associated with a product implementation from purchase to deployment and finally decommission. Also known as fully-burdened costs, after it is determined, TCO is a cold, hard kernel of financial truth that reflects the true financial impact of the resources tied up in an environment.

Here is a brief example. The Goodrich IT department decides that to justify decommissioning a duplicate development environment, they need to provide a TCO metric that demonstrates the significant costs associated with the environment.

The team does some fairly intensive research that produces the following data. Keep in mind that the capital costs for the storage frame and the tape library are depreciated:

The 3.5 TB of storage attached to the frame was purchased two years ago for $350,000 (the frame still has one year of depreciation remaining).

The frame has an annual hardware and software maintenance contract of $100,000.

The datacenter support team has analyzed the cost for providing and supporting one square foot tile of datacenter space, and it determined the cost to be $10,000 per year.

The footprint of the frame itself is 1.5 feet and the frame consumes 1/100 of the power and cooling cost of the datacenter, which comes to $50,000 per month.

The development environment is backed up once a week to its own dedicated tape library, which was purchased at the same time as the frame for $250,000.

Goodrich pays $50,000 a year for tape cartridges and offsite storage of tapes for this datacenter. This development environment consumes 1/25 of this resource.

Neither the frame nor the library is attached to a SAN. The frame and the library are only 60 percent allocated with all ports in use.

The storage support team is comprised of two full-time employees; each earns $75,000 annually. They spend 50 percent of their time supporting this development environment.

Chapter 1, are not easy figures to come by. Analysis of the environment from the application level to the physical allocation level (using SRM tools and homegrown scripts) is required to get a true picture of both allocation utilization and utilization efficiency. These numbers, when taken together, provide a better picture of the storage TCO. For the purposes of this example, if the allocation rate is 60 percent, but the utilization efficiency is only 30 percent, the annual TCO for the storage in this environment is actually $0.37, or almost quadruple the cost estimates from the application team!

One other caveat to remember when determining the TCO for storage in an enterprise is that costs fluctuate over time and that solutions at different tiers have different cost structures. Obviously storage costs have decreased drastically over time, but the costs of labor and facilities also vary. Most costs typically decline over time and there needs to be some consideration for the decreasing costs, especially if the IT department or storage infrastructure teams plan to utilize a cost recovery or charge-back model for funding services. Fully burdened costs are required to facilitate the move to alternative funding models and the move to tiered storage; however, these costs must be accurate to avoid both overcharging the client as well as ending up short of funds.

Even though it can be painful, the math behind TCO is relatively simple, and although this example is a simplified overview of the elements involved with calculating TCO, TCO analysis is on the whole relatively straightforward.

Now is probably a good time to take a quick breather before going back for more. Thus far, I have given a brief overview of storage networking technologies and their most common usage scenarios. In addition, I broached the topic of TCO.

Now armed with both an overview of the storage networking technologies and a cursory understanding of how TCO analysis works, we can move on to the spicier materials behind financial metrics in Chapter 3.