Server performance and benchmark testing guide
A comprehensive collection of articles, videos and more, hand-picked by our editors
This tip is part two of two, originally published as part of "Tuning Performance and Capacity Management," chapter...
two of the Choosing Performance Monitoring Tools e-book. Part one reviews performance monitoring tool features.
In complex modern data centers, performance monitoring is more important than ever, but finding the right server monitoring tool for the job can be a challenge.
Businesses rely on performance monitoring tools to ensure business productivity via application performance and uptime. Your server performance monitoring tool delivers important metrics on server and application performance levels, service levels, and even network issues that create bottlenecks. It also provides a centralized view of the data center's physical servers and virtual machines as well as devices and applications, enabling IT professionals to proactively address failures and improve the user experience.
But not all server monitoring tools are created equal. Some may have trouble measuring physical and virtual environments together; some may provide too much information or information that is too granular for data center managers to put to good use. Still others may not be cost-effective or provide only stripped-down functionality. It's important for IT procurement managers to explore the range of tools and make informed decisions about which server monitor is right for their environment.
With the wide range of offerings and functionalities, choosing the best server performance monitoring tool can be a challenge. These four steps can lead the way to the right selection:
- Analyze your environment. Start by analyzing your applications and the environment in which you run them. Most organizations have a significant on-premises footprint, so the ability to monitor hardware is important. On the other hand, public cloud computing and complex applications will be a part of every IT organization's future (if not its present), so defining requirements in light of extended functionality options also matters. Create a list of current and future performance monitoring requirements and use it to filter potential vendors.
- Define your budget. Your tool options range from free to expensive, with functionality and ease of use likely to improve as costs increase. Examine your budget and estimate what you can invest in performance monitoring. Factor in the cost of any application downtime that could occur without the right server monitoring tool. That calculation may free up some additional money.
- Choose your favored deployment option. In the past, Software as a Service (SaaS) performance monitoring technologies offered less functionality than their on-premises counterparts. Today, that isn't necessarily true, so the deployment choice comes down to preference. Choosing one deployment option over another will filter out some products and simplify the decision-making process.
- Create a shortlist and perform a pilot. It's almost a given that all products perform well in the vendor's tests and in demo situations, but some will operate differently in the real data center. Test-drive your top three choices. The tools will be difficult to fully deploy; instead, create a scaled-down application and apply each of your shortlisted performance monitoring tools to it. This will help you understand how they work and, critically, compare the monitoring tools' functionality and ease of use.
As you move forward in the selection process, there are a couple of "gotchas" to consider:
- Future-proof your selection. Most companies will use public cloud computing, so any product you consider should fully support that transition, even if it isn't on your short-term roadmap. Consider both today's and tomorrow's infrastructure environment.
- Understand what you are paying for. With so much competition in the performance-monitoring space, vendors may offer an introductory feature set at rock-bottom prices, holding back important functionality as part of pricey add-on modules. If you expect the full product suite, don't budget for a base product with extended functionality only available with an additional purchase. For a SaaS server monitoring tool, understand the factors that affect pricing: the number of applications monitored, the number of user accounts registered, the number of components being connected and so forth.
- Prepare for employee training. Even if a vendor strives to make its product easy to use, the reality is that complex applications require more complex monitoring tools. Your staff will encounter a learning curve with whatever performance monitoring product you select. Invest in training and you'll reap the benefits of monitoring as quickly as possible.
About the author:
Bernard Golden is the former CEO of HyperStratus, a cloud computing consultancy. He is also the author of four books on virtualization and cloud computing. Golden is a highly regarded speaker and keynotes conferences throughout the world.