Best Practices for Improving Network Performance
Networks have changed. Where once most networks were simply a means to end (i.e., a way to connect critical components of an IT environment together), today they’ve become far more complex and play an increasingly crucial role in an organization’s ability to serve customers.
Obviously, different size organizations and industries experience different challenges with their networks (for example, high-frequency trading companies consider latency before almost anything else, SCADA shops put a premium on reliability and stealth technologies, and some other industries put ease of use first). But, in this new digital age, nearly all companies are looking for ways to improve and optimize their network performance.
Below are some best practices for doing that:
1) Simplify Your Network Design- Many organizations have either inherited an outdated network design or discovered a manufacturer did what was best for the manufacturer instead of their organization. Unfortunately, a bad or poorly designed network, or even a network that isn’t optimized, can create a lot of complexity.
If your organization is in this situation, take time to review your network design and simplify and standardize it as much as possible. Simplifying your design will allow a new employee or consultant to easily look at your network, understand it and fix it, if needed. (While new ideas from manufacturers can be highly enticing and seem like they will simplify your life, it’s important to do a cost benefit analysis and ask questions such as: If a single switch fails what happens? If a single fabric that is the entire six switch three way redundant core fails then what?) Standardizing your network will also result in less work when upgrading, changing or troubleshooting. The point is: Don’t do work twice where once will do. Along with this, be sure to standardize your operating systems, products and configurations so they work the same across your entire environment. That way, if your network breaks or comes to a halt, you can quickly figure out where the problem is and fix it without having to consider multiple dependencies. (Note: Several tools and platforms are now available that can help extrapolate the management layer and maintain standard configurations across a heterogeneous multi-vendor environment.)
2) Improve Your Visibility and Monitoring- Most networks slow down at some point. If you don’t have the right equipment in place you won’t have any visibility into nor any way to identify problems when they happen or more importantly, before they happen. Having visibility is especially important whenever there are multiple system problems involved, as it can often be difficult to tell: 1) which network system (internal or external) or application is causing the problem; 2) whether it’s a catastrophic or simple issue; or 3) where and how to fix it. A great way to gain this visibility is by having a single management interface for all technology and functionality. Then, whenever bottlenecks happen, you can solve the problem while continuing to maintain full network performance.
3) Figure Out What Your Organization’s Current Network Performance Is- The key to any network performance is having a system that provides a service level agreement (SLA) in real-time. Before you can do that though, you need to know: What is your organization’s current baseline for network performance? Once this baseline is established, the system can then monitor your network performance, alert you to any changes from that baseline and tell you where the variances are.
4) Determine What Your Organization Expects the Network Performance to Be- After you know what your organization’s current network performance is, you then need to know: What does your organization expect it to be and why? This information will serve as your North Star in improving your network performance.
5) Pre-model Changes- Many organizations are reluctant to make changes to their network or operating system or add any new technology for fear they’ll disrupt their network or users, and/or cause a network outage. While this is a legitimate concern, many new technologies are now available that allow you to pre-model, test and review any changes you want to make to your network—before you actually make them. Then, if there are any issues, these technologies will tell you exactly where and how to trouble shoot them. That said, if you buy this technology, be sure to look for a system that’s based on open standards vs. a closed-loop system so you’ll continue to have flexibility as your network grows.
Moreover, if, as part of your network performance optimization practice, you decide to add tools, don’t forget to allocate appropriate resources to stand them up and maintain them. Otherwise, they’ll just be clutter and make your life more difficult in the long run. Plus, your IT or financial leadership will question why you didn’t know something or why you had to bring in a consultant, etc.
By following the best practices listed above, not only will your organization have a faster and more efficient network, but you’ll also have one that’s more resilient.