The Impact of Monitoring on Network Scalability

By Leon Adato, Head Geek, SolarWinds

IT professionals are constantly planning for growth—it’s ingrained into our very being. But today, it’s more complicated than ever before as enterprise network architectures have evolved from flat networks where everything was interconnected to hierarchical models with enhanced security. We are now operating in borderless environments, and planning for growth (i.e. achieving network scalability) in this brave new world isn’t as simple as it once was.

Despite the changing network landscape, one constant remains in terms of building an effective network scalability strategy—monitoring. To best underscore why monitoring is so crucial in achieving network scalability, let’s first explore the current drivers of network complexity.

Today’s Network

Unless you’ve been living under a rock, you’re well aware of BYOA (bring-your-own-anything), but it’s worth exploring its modern offshoot, the Internet of Things (IoT), as well as other trends like hybrid IT and software defined networking (SDN) to understand the effect each is having on today’s networks and today’s network administrators.

IoT is now a consideration worth separating from the overall impact of BYOA. Network engineers should be thinking about these two technologies disparately due to the sheer volume and variety of devices being connected to the network by employees—think fitness wearables, laptops, tablets and smartphones for starters. These items should be considered when planning for network capacity as well as network policy—should they be allowed on the network due to security and capacity concerns? Many companies, after truly understanding the scope of these devices’ impact on the network are now allowing their employees to bring these devices on the condition that they connect to a guest network to prevent potential security and bandwidth issues.

IoT, however, is different from BYOA in that the data on the network is part of the larger corporate strategy, whether they fall under the category of whimsical, such as coffeemakers and toasters, or more realistic needs, such as heating and ventilation controls or tracking the location of office systems and other business resources.

Next, the existence of the cloud and subsequent transitioning of some infrastructure offsite while some remains on-premises has led to a hybrid IT reality for many organizations. “But the cloud doesn’t have anything to do with our network,” you say? Think again. The cloud has a major impact on network bandwidth. And planning for the bandwidth needs of known cloud services is difficult enough, but even more difficult is planning for the unknown—end users leveraging online storage, file sharing and other cloud-based services without IT’s knowledge. This, of course, has security repercussions, too—free or freemium cloud services are becoming just as much an aspect of the shadow IT phenomenon as anything else.

Lastly, SDN. No, despite what SDN vendors may tell you, SDN’s effect is minimal for now. It’s primarily a concern for large carriers, such as AT&T, Time Warner or Cox, but it’s not on the radar at the WAN connection (and certainly not at the LAN connection). So, ultimately, my recommendation is to be thinking about it and start planning, but in action, SDN should not be a priority just yet.

How Network Monitoring Can Help

By understanding the modern network landscape, it’s easier to understand the context of network scalability concerns, and how network monitoring is a critical piece of that, not just an afterthought. The key to effective scalability is setting expectations for what the network can handle, what it can’t and how it will be supported.

As an example, if an organization adds 3,000 IoT-related devices to the network, managers need to know what the threshold is before the network goes beyond capacity. In other words, if 3,000 devices turns into 4,000, the network may not scale.

At a high level, monitoring is part and parcel of the entire network capacity process. Specifically, it’s one thing to project growth, but monitoring provides the bridge—the cold, hard data—between reality and projections. Any capacity planning that does not use network monitoring as a foundation to build upon is incomplete.

Monitoring is such an important part of building a scalable network that it may be worth considering monitoring as a discipline, which varies from basic monitoring in that it is an actual role, an assigned focus of one or more individuals within an organization. Such a role provides immense value through the ability to turn disparate data points from various monitoring tools and utilities into more actionable insights. It considers all, and does so from a holistic vantage point.

Best Practices

So as BYOA, IoT, and hybrid IT (and, at some point in the future, SDN) become more ingrained into our networks, monitoring, whether done by any number of team members fulfilling multiple roles or ideally, as a discipline, is a simple, effective way to manage the breadth of the growing network environment. Whether you’re scrambling to solve a scalability problem after it’s happened, are proactively planning for growth or need to replace your monitoring solution altogether, consider employing the following best practices to move the organization forward with as little friction as possible.

– Adding a dedicated monitoring expert or experts who can provide a holistic view of the organization’s infrastructure performance, turning seemingly disparate data points gathered by monitoring tools into valuable, actionable insights. Work with your organization’s existing business growth projections to calculate what your experts will need.

– If a dedicated expert is not possible, ensure the current IT team understands the nuances of monitoring hardware, networks, applications, virtualization and configurations and has a comprehensive suite of monitoring tools available.

– Ensure your team’s chosen tool or suite of tools has elements that address your environment’s scalability and implement the solution right away.

IN conclusion, it’s also important to remember that growth and capacity planning are not challenges limited to the network. In fact, as the line between network and systems continues to blur with greater convergence, it’s more critical than ever that monitoring capabilities extend beyond traditional silos and provide visibility across the entire stack.