Server performance and benchmark testing guide

Today’s servers are expensive, high performance machines. So, getting the most out of those machines and keeping them running smoothly is critical to maximizing their value. However, maintaining high levels of server performance and keeping up with growing computing demands doesn’t happen on its own. Tracking performance and making the most of hardware resources is an ongoing process that can keep data center administrators busy. But, there are ways IT professionals can simplify server management while maintaining top performance. Benchmark testing can help administrators monitor server resources, manage utilization, optimize performance and even identify problems before they occur. Simple server configuration tweaks can squeeze more performance out of both physical and virtual machines (VMs). This guide will cover best practices for benchmark testing and includes valuable tips from our data center experts on how to improve server performance.

Table of contents:

Tools and strategies for benchmark testing

Benchmark testing can help establish baseline performance expectations and track a system’s performance over time. Administrators can also use benchmarks to help with virtualization capacity planning and better manage server resources, utilization and performance. This can allow data center administrators to identify potential performance problems and plan for future capacity needs. In this section, we offer several tips explaining how to pick the best benchmark testing applications for your organization, how to get the most out of the tools you choose, explain what metrics are important and help you evaluate benchmark results.

Selecting benchmark tools for specific needs
The good news is that there are a plethora of server benchmarking tools available to administrators. The bad news is that not every tool will fit an organization’s benchmark testing needs. Most server benchmarking tools will address basic tasks, but choosing the “best” benchmarking tool can be a confusing process. The key is to look beyond the basics and decide which tool will work best with your organization’s hardware and ask yourself if it can provide the level of technical detail needed to track the performance metrics important to your environment.

Using benchmarks for virtualization capacity planning
Ensuring a virtualized data center has the computing resources to keep up with future growth is critical to maintaining a competitive business advantage, but virtualization capacity planning doesn’t happen on its own. It takes time and a firm understanding of how business goals align with hardware resources. Benchmark testing can help solidify that understanding by predicting increases in capacity, stabilizing user performance and even identifying potential problems before they occur. By using, and evaluating benchmark tests, administrators can effectively plan their infrastructures before resources become depleted.

What to look for when conducting benchmark tests with PerfMon

Microsoft's Performance Monitor is a powerful tool that can conduct benchmark tests measuring a wide range of processes and server aspects. However, the variety of settings and counters available can make interpreting the results difficult, and all that information can distract administrators from the important information buried within the numbers. It is not always easy to know what to look for when conducting benchmark tests. In this tip and expert lists the types of metrics that administrators are likely to see with PerfMon and explains which are the most important and how to interpret the results.

Server benchmarking tools and stress testing

Using just one tool to perform benchmark tests sometimes doesn’t provide accurate data. Administrators should turn to several benchmarking tools and then compare results to verify accuracy. This tip outlines the importance of stress testing and lists several tools that IT professionals can use. Some benchmark testing tools are designed for physical hardware assessments and are not suited to gather accurate data in virtualized data centers. In this tip, an expert lists a handful of benchmarking tools and explains their strengths and weaknesses.

Benchmark testing strategies and best practices

Benchmark testing can be a powerful tool for data center administrators, but benchmarking must be used appropriately. When conducted properly and consistently, good benchmark testing can provide valuable information that admins can use to improve server performance. However, even when benchmark tests are successful, understanding and using the resulting data can be confusing. Results can be misinterpreted and hard to understand. In this section, we offer tips that explain benchmarking best practices that can reduce potential errors and help administrators gather more accurate results.

Recognizing and avoiding common benchmark mistakes
Using server benchmark testing is critical to maintaining long-term system performance. However, those benchmark results will only be helpful if administrators follow best practices to collect relevant information. Looking at the wrong metrics, such as overlooking network bottlenecks that may not show up at the server level, can make benchmark testing inadequate, or even counterproductive. These benchmarking best practices can help administrators avoid common benchmarking mistakes and know how to use benchmarking tools the get the right results.

Best practices for server benchmark testing
All metric testing will produce data, but when best practices are not followed, that data may not be relevant or helpful. Following a structured approach to benchmark testing will ensure the correct metrics are being measured, variables are accounted for and the results are accurate. In this tip, an expert explains his six-step process for server performance testing that can help administrators reduce errors and get more relevant results.

How should benchmarks be used to optimize virtual workload balancing?
Although virtualization comes with undeniable benefits, it can also represent a serious drain on computing resources in a data center. Unfortunately, as virtualization has become more widespread, performance monitoring has not been applied to VMs. As the number of VMs in a virtual environment grows, they must be monitored against an organization’s benchmarks. Using benchmarks and tools for virtual workload balancing will help you plan your next VM deployment.

Optimizing server performance

There are few steps an organization can take to improve data center capacity and performance without spending large amounts of money. However, fine-tuning server configurations and eliminating bottlenecks can lead to substantial performance gains with minimal expense. Ensuring that systems are set up correctly is especially important when talking about a virtualized infrastructure, where CPU contention or I/O bottlenecks can lead to serious problems. Often times, small changes to a server’s BIOS settings can make a big difference. But, administrators should consider these performance options carefully because they sometimes come with trade-offs, such as cooling problems or increased energy consumption. In this section, we offer a selection of tips explaining how to improve server performance while maximizing return on investment.

Spotting undervalued server features
Server’s aren’t always at their best right out of the box, and setting up a new server is more complicated than just powering it up. Administrators must now consider how the server fits into the overall environment and then adjust several configuration settings. While much of this process is intuitive, there are some lesser-known features that can boost server performance. But before you begin adjusting those settings, be sure you understand the implications, which could include overheating or even reduced performance.

Monitoring and optimizing host server performance
Choosing the right hardware, and then tuning that hardware, is the backbone of successful visualization deployments. Optimizing host server performance is the first step in ensuring VMs have access to the resources they need to run smoothly. For example, having the host operating system on a dedicated drive prevents it from having to compete with VMs for disk resources. In this tip, you’ll also learn how to choose a host operating system and how to balance physical CPUs with virtual CPUs.

Tweaking BIOS settings to improve system performance
By adjusting BIOS settings, administrators can ensure their servers are running at peak performance. Optimizing these settings can lead to significant gains in computing power, but be aware that changing some BIOS settings could also compromise system stability or augment overheating problems. In this tip, an expert explains which BIOS settings can safely be adjusted to provide a noticeable improvement in server performance.

Use Nagios to trend and troubleshoot performance issues
Data center administrators can’t keep an eye on server performance 24/7, but they should have a way of tracking performance that can notify them of potential problems. While Nagios won’t improve server performance on its own, the data it gathers can help administrators identify and track performance problems on your networks. In this tip, an expert writes about how Nagios can identify performance problems and explains how the tool’s reports can be used to improve performance.