Measuring network complexity and why it is essential to business

IT networks serve as the circulatory system of businesses around the world, but the rise of the BYOD culture, the need for 24/7 uptime, and a move towards virtualisation and the cloud has seen them wrestling with increasing complexity.

If this growing complexity isn't properly managed, it could have a significant impact on the way businesses function as well as their balance sheets and brand reputation.

For 30 years, the network has been perceived by management simply as a form of plumbing - but the network is much more important than just a mix of switches, routers and load balancers. It's a collection of people, data, devices and infrastructure – the endpoints that all interact in order to make things happen.

From a business' point of view, the complexity of its network can be judged by the number of interconnected endpoints, the number of connections between these endpoints, and the way information moves through these connections to perform business processes.

Essentially, the more endpoints that support a business, and the more critical business processes that utilise those endpoints, the higher the level of complexity will be.

What does network complexity mean to a business?

As a business' customers become ever more connected, its network may need to grow too – possibly exponentially – as it becomes ever more entwined within a widening circle of users, devices, sensors and apps.

If this proves to be the case, then the biggest impact to business networking is likely to be the cost of managing this growth. The emerging risk is that the rising cost of managing network complexity might one day outweigh the value.

When small, a network can usually be managed through manual processes with no significant impact from mistakes - but, as it grows, and users, devices, sensors and apps are added, greater inefficiencies and vulnerabilities can develop. If network complexity is allowed to increase without sufficient checks and measures in place, it could have dire consequences.

In the worst possible situations, a network failure could see electric power plants going offline, planes remaining on runways at airports, consumers unable to withdraw cash from ATMs, and McDonald's unable to serve Big Macs and fries.

Measuring network complexity

When monitored and managed, continuing growth in network complexity needn't result in such negative outcomes. Indeed, Internet giants such as Google and Amazon are able to successfully manage an ever-increasing degree of demand and resulting complexity, increasing the reach and scope of their organisations' networks, while maintaining relative constancy in cost.

As a result, I'd argue that a key factor in successfully managing the growth of a network's complexity is the ability to measure the rate of that growth, by quantifying the complexity level at any given moment in time.

Measuring a network's complexity can raise awareness of a situation which, although identified, is not yet fully understood. An insight into its own network's complexity will provide a business with a sense of perspective, enabling it to more efficiently and appropriately manage network growth and eventually include the full network – not just networking hardware – as an entry on balance sheets.

Such a measure can be used as a strategic guide post, allowing businesses to focus on taking the steps necessary to drive costs down – or at least keep them constant – while ensuring that they are able to efficiently carry out core business functions as their network complexity increases.

What's the solution?

The ultimate aim of quantifying a business' network complexity is to better understand the change in the level of reliance that a business has on its network: how intertwined the network is with all of the core functions of the business.

Now a new open-source software application, Tapestry, freely available as a download from FlowForwarding.org, has been designed to capture data from a network's DNS servers and process it through an embedded equation to generate a Network Complexity Index (NCI).

By taking regular measurements, a business can gain information on just how fast its network complexity is growing. An increased focus on the speed at which this growth takes place is needed to allow network managers and IT directors to better manage it. Measuring NCI can provide additional data to allow them to consider the cost implications and the available solutions – not only to keep these costs down, but also to avoid the network becoming overwhelmed by the functions that inherently rely on it.

Reducing the cost of complexity with technology

Following years of viewing the network as plumbing, knowing a network's NCI should encourage IT professionals to see the network differently, providing them with a new perspective on solutions available today such as automation, and those of the future, namely software-defined networking (SDN).

Through automated network management, for example, a business can better maintain control of its network infrastructure.

By making use of sophisticated, affordable software and data analytics within the control plane, automation allows network managers to carry out more valuable tasks, offloading their mundane day-to-day actions and focusing on those which can add value to their organisation. This, in turn, immediately makes the IT operation more efficient and cost-effective. In this way, automation can help businesses cope with growing complexity and enable them to scale while keeping costs down, even as the network grows.

SDN is an especially hot topic in the networking space right now. By deploying commodity hardware controlled centrally through an intelligent software layer, SDN introduces programmability to the network.

If you consider the majority of the existing networking landscape as hardware-defined networking (HDN), an entire industry organised along box function lines – switches, routers, load balancers, and the like – you'll notice that, under the hood, the boxes are actually much the same.

SDN allows companies to use open standards to program remotely in real time and turn a single cost-effective white box device into a switch, a firewall, a router or a load balancer. White boxes can be quickly reprogrammed to perform some other role as needed.

It's hard to predict when SDN will break into the mainstream but, when it does, it has the potential to reduce the cost associated with the natural rise in network complexity in modern organisations. The market correction that SDN promises could occur almost overnight and those businesses which have planned for its arrival are the ones that will gain a competitive edge as they break through the network cost-complexity barrier. Calculating the network's complexity should be the first step in that plan.

Into the future

As more users join a network, with more devices, each with more sensors and apps, and all sharing more information with each other, so the network's complexity increases naturally.

And as complexity grows, so too does the cost of managing this complexity. And as more core business functions become entwined within the network, there's a growing risk of it not being able to deliver as required or even being overwhelmed entirely. For the first time in history, it is becoming clear that complexity rather than bandwidth is the barrier to network growth and scale.

Measuring and managing complexity has potential to help businesses run highly scalable, adaptable, automated, software-defined networks.

The future belongs to the organisations and people that embrace the complexity and the value that the real network has to offer.

Stu Bailey is the founder and CTO of Infoblox.

Image: Flickr (luc.viatour; Skokie Public Library; Cea.)