The Three Capabilities Needed for Network Performance Management of Virtual Server Infrastructures
by January 8, 2013

Filed under: Network Performance Management, Performance Management Tech

server virtualizationServer virtualization is one of the most important IT trends of the past decade. According to Gartner, server virtualization has recently surpassed 50% of all server workloads, and the number of virtualization workloads is likely to grow fivefold by 2015. Drivers for this rapid adoption include reduced hardware costs, improved resource utilization, radically streamlined provisioning and more – all leading to improvements in data center efficiency and lower IT costs.

However, virtualization places a greater strain on network performance as more real time applications share a single network interface. Even with a 10Gig interface this creates a bottleneck, which can affect end users’ application performance experiences.

As IT organizations embrace server virtualization, they must pay greater attention to network performance management and performance tuning. Virtualization changes network dynamics, and the ability to collect and analyze the relevant data is essential to managing network performance across both physical and virtual infrastructure.

Traditional Network Performance Management solutions aren’t designed for the virtual layer that now exists. Instead, network engineers need a next generation solution that understands the end-to-end experience all the way into the virtual environment in addition to the physical network.

Network engineers need three capabilities to monitor and manage network performance across virtual server infrastructures:

  1. A solid baseline understanding of the network’s behavior over time, including achievable bandwidth all the way to the virtual machine.
  2. Real-time network visibility end-to-end from the virtual infrastructure to the end user (even if that user is “remote” from the LAN at the main office).
  3. Alerts and automated reporting on network performance with application specific performance thresholds through the virtual infrastructure – to drive proactive troubleshooting and problem resolution.

Instead of using multiple tools to achieve these capabilities, look for a single solution that can achieve all three. When researching the appropriate NPM tool, make sure you consider as many factors as possible.

What issues have you run into while trying to manage the network performance of virtual server infrastructures?