Capacity planning makes it possible to know if your current infrastructure is adequate to cope with the addition of new applications, or understand the impact of greatly increased traffic volumes, and to determine how much extra equipment may need to be deployed if more resources are indicated. But how do you select the best capacity planning tool for ensuring data center efficiency?

If you are commissioned with finding the best tool for optimally configuring systems within your IT organization or the overall efficiency of your data center, you have a real challenge. Many companies offer “capacity planning” tools, yet it isn’t obvious based on their marketing literature which ones are really worth considering.

Anybody with a dartboard can claim to have a server capacity planning tool. Unfortunately, companies selling dartboards for capacity planning aren’t likely to be very honest about the sophistication of their tools. The best advice is caveat emptor - let the buyer beware.

So how do you avoid buying a dartboard when accurate predictions are what you really need? What makes a good capacity planning tool? What should you look for in a capacity planning tool? To answer those questions, let’s start by discussing what is NOT considered to be good enough for capacity planning.

What Is NOT Good EnoughPerformance Monitoring

First, performance monitoring is not capacity planning. Receiving an alarm event 15 minutes before users complain is not capacity planning. Monitoring and alarm alerts are essential components for capacity management, but they aren’t considered capacity planning.

Graphing Historical Data

Real-time performance data isn’t enough either because capacity planning always requires a historical record of some sort. And the practice of keeping some historical data available for graphing and charting is insufficient. You need more data beyond real-time to conduct accurate capacity planning.

Trending

Beware of companies that promote trending as capacity planning. Computer system performance is not linear, and a capacity planning tool needs to analyze more than just past system performance to make accurate predictions about the future. Trending is of no use for many projects that involve capacity planning. For example, you can’t plan a server consolidation project with trending.

Simple Math

You should also be wary of tools that claim to do capacity planning for server consolidation by simply adding together the resource utilization of each of the workloads. After normalizing CPU utilization to account for differences in computing capability, the utilization for each workload is added together to determine how much of the target CPU will be utilized after consolidation. A similar calculation is performed for other resources such as memory, I/O, and the network.

This kind of simplistic procedure can be deemed good enough to find potential consolidation candidates, but it leaves way too much out of the equation for making the final decision when consolidating important workloads. You need a tool that truly understands the details regarding your server architecture, your applications’ use of that architecture, and how workloads will interact when they are consolidated.

What You Really Need: Modeling

Much has been suggested about what isn’t a good capacity planning tool. Now, you’re probably wondering what a good capacity planning tool actually is. If you are serious about capacity planning, especially if critical applications are involved, you need a capacity planning tool that includes a strong modeling component.

Sometimes when people talk about “modeling“ they mean a description or diagram. That’s not the kind of model we are describing here. You certainly need a description of the systems involved, but that description is really just one step in a good capacity planning process. You need a modeling tool that can look at that description, analyze the incoming workload information, and predict how the systems will perform.

Simulation Modeling

Consider two types of modeling methods to predict performance: simulation modeling and analytic modeling. A good simulation modeling tool will create a queuing network based on the system being modeled and pretend to run the incoming workloads on that network. Simulations like these can be very accurate, but a lot of work is necessary to adequately describe the systems with enough detail for the results to be dependable.

Simple Queuing Network

More Efficient: Analytic Modeling

Analytic modeling also takes queuing into account, without pretending to run the incoming workloads on the model. Instead, in a good analytic modeling tool, formulas based on queuing theory are used to mathematically calculate processing times and delays. This type of modeling is much faster and not nearly as tedious to establish. And the results can be just as accurate as with simulation modeling.

Analytic models are not as generalized as simulation modeling. When a crucial situation arises where a suitable analytic model is not available, it makes sense to utilize a simulation model instead. The rest of the time you should use a much easier and faster analytic modeling process.

Analytic modeling is usually what you want in a capacity planning tool. If you want to cover your bases, acquire a tool that performs both analytic and simulation modeling. Be sure to check whether the vendor isn’t misusing the term “analytic” when making claims about their tool. Make sure that your tool selection uses sound methods based on queuing theory to make its calculations, not something more closely resembling the less accurate capacity planning techniques described earlier in this article.

Ongoing Process

Finally, after selecting your capacity planning tool you should commit to an ongoing capacity planning process, taking into account growth rates and changes in your IT environment. Further, if your organization has to deal with huge traffic volumes or rapidly fluctuating loads, interim capacity plans should be done at least every three months. In addition, ad-hoc reports should be generated to verify all is well or to investigate unusual occurrences and follow up on unusual patterns.

Follow a process that forces you to:

• Understand business objectives

• Prioritize services and risk levels

• Establish service levels

• Plan and provision services

• Manage service performance

• Track performance against service levels

This procedure should be carried out in tandem with ongoing performance management to detect and prevent bottlenecks from materially affecting system performance. Executing the ongoing process of capacity planning will result in increased efficiencies within your data center.