Setting up a stable and robust networking environment is essential to running a business in today’s marketplace. But, like with so many things, your environment won’t stay this way forever – that’s where the importance of monitoring, network performance testing and baselining come in. For your network to run as efficiently as possible at all times, it’s essential to understand the ins and outs of your environment well enough to be able to reliably predict how your network will be operating at any given time. In this blog, we’ll look at the value of proper monitoring, and why baselining should be the backbone of a network performance test.
Network Performance Baselines make it easy to tell when your network is over or under capacity.
In order to properly measure the performance of your network, you need to understand how it performs when all conditions are ‘normal’ – in other words, you need to monitor, test and report on your network’s connectivity, average bandwidth usage, protocol usage and peak bandwidth usage over a period of time. Once you’ve established the average behavior of your network environment, you can use this baseline as a guide to alert you to any potential problems within your environment.
Baselining indicates when resources are at capacity and a network upgrade might be necessary.
Because baselining makes it possible to measure the current network utilization against your base utilization, you’ll be able to see when the average starts approaching capacity. This makes it invaluable in planning for upgrades, since it can be used as an ‘early warning system’ for network capacity and service issues. Rather than running out of bandwidth and letting your organisation suffer while you implement upgrades, changes to your environment can be made incrementally and in a way that doesn’t disrupt workflow.
Baselines and Service Level Agreements go together like 0’s and 1’s
Setting a good baseline is also invaluable in setting Service Level Agreements (SLAs) that are accurate and beneficial to both you and your customer. Your baseline should already provide you with a good idea of the capacity restrictions of your environment, which will allow you to set realistic SLAs that are within the scope of your ability to deliver. Using a Netflow Analyzer is useful in identifying exactly where the bottlenecks and service interruptions are happening on your network, so it’s vital to use some form of Netflow analyzer on your network.
Consider virtual as well as physical devices
Network virtualization is increasingly common in today’s environments. Without proper management, this can present a problem with regard to network discovery and performance testing, so it’s imperative to be aware of the virtual switches and devices that exist within your environment. Keeping communication channels open between all members of your IT staff is essential in avoiding virtual devices that fly under the radar of your monitoring and performance testing tools.
Baselines aren’t static and should always be open to revision and optimization
Markets today are much more volatile than a few decades ago, and this is especially true for any organization with even a peripheral involvement in Information and Communications Technology (ICT). For this reason, it’s of paramount importance to continually revise and optimise your baseline through network performance tests. For Wide Area Networks (WANs) that span large physical distances, as well as networks that rely on many virtual devices, this is even more of a concern – if physical or virtual devices are added to the environment without the administrator’s knowledge, being lax with network performance testing could have dire consequences. Performing regular audits of devices and WAN and internet links, as well as physically monitoring your environment where possible is a good way to stay ahead of changes to your network environment.
To find out more about how IRIS can assist you with creating a network that works for the unique context of your organization, download our free Guide to the Effective Management of a Large-Scale WAN.