The up and coming generation of network performance monitoring, what to expect


The size, scope, and complexity of networks have increased significantly in the last few years. Now it’s more important than ever for companies to ensure their networks run reliably, so the pressure is rising for Infrastructure and Operations (I&O) and IT teams. 

As a result, everyone is looking for ways to improve network performance monitoring. Enterprise IT teams are watching each other to see what tactics they adopt to manage their networks more efficiently and increase network performance. 

This in turn is driving growing adoption of network monitoring tools and network management systems (NMS). According to a recent report, the NMS market is expected to grow at a compound annual growth rate (CAGR) of 9.4%, to reach over $14.6 billion by the end of 2027. 

But these figures mask a number of different factors and trends. What can we expect for the up and coming generation of network monitoring? 

1. AI networking

AI networking, which also encompasses self-healing networks, is a subset of AIOps which takes automation a step further. It uses artificial intelligence (AI) to automate many routine tasks involved in maintaining and enhancing network performance. For example, AI can automate troubleshooting, automatically set baselines for smarter monitoring and alerts. According to a recent Gartner report, AI networking can reduce certain network operational management costs by up to 50%.

With self-healing networks, AI can assist NetOps teams by automatically applying actions to manage Day 2 network operations more efficiently. This has AI network performance tools moving beyond delivering alerts or making suggestions for next steps, to actually making and acting upon decisions. 

For example, wireless access points can detect increased noise, congestion, or interference, and reroute to an alternative channel. Networks can respond immediately to spikes in demand by dynamically requesting more bandwidth or rerouting traffic. 

Gartner places AI networking use cases like automated incident response and network automation prominently on the Slope of Enlightenment for its 2023 Hype Cycle for I&O automation, estimating that it will take less than two years for the former and up to 5 years for the latter to reach a plateau. 

2. Automation

As networks become more complex and sprawling, encompassing more devices and switches, network managers are turning to automation to keep on top of the workload. Intent-based networking (IBN) automatically adjusts and optimizes networks, configures settings, and creates policies. This results in fewer errors and faster response times – for example, configuring security settings manually for an extensive business network can take hours if not days – and thus increases security and uptime.

Gartner calculates that approximately two-thirds of network tasks are currently manual, and that most network vendors have technology that addresses a significant portion of them. With networking adoption estimated at under 10%, this leaves room for massive uptake in the near future. 

Automation can also make sense of network data to produce insights into network behaviors that were previously only guessed at. IT and network ops teams can use these insights to anticipate performance issues and prevent outages. 

3. Smart security

Our increasingly complex networks are also increasingly vulnerable. Network management teams need to be certain that no rogue devices or users have entered their sprawling networks. They also need to deal with the issue of employees using devices that aren’t part of the LAN, like their personal smartphones, for work purposes. These devices could access and transmit vital business data, so network monitoring needs to extend to include them in all security planning. 

In the face of these challenges, we can expect to see more adoption of zero trust security frameworks to protect Network Performance Management solutions, since they have access to all critical infrastructure. More teams will employ AI to monitor metrics, logs, and events, driving up observability to enable greater visibility into network activity across networks, devices, and IoT systems. 

Predictive analytics is not yet feasible for network management, since it’s not really possible to predict security issues before they happen. However, netops teams are starting to use AI analysis and smart correlation for early issue detection, enabling them to spot existing issues faster, while they are still minor. This way they can resolve them before they grow into serious incidents, helping harden the network’s security profile. AI analytics enables IT teams to connect the dots across disparate networks to draw more accurate and informative conclusions about network security levels, vulnerabilities, and what to do to address them. 

4. Open source

Many AI projects are based on open-source protocols, and that’s likely to increase in the near future. The widely-used SNMP (Simple Network Management Protocol) is itself an open-source network management protocol that serves as the foundation stone of network performance monitoring. 

We’re seeing more and more open-source projects like OpenConfig/gNMI and OpenTelemetry, which could bring new approaches and add more depth to network monitoring. AI-powered telemetry could improve insights into network performance. 

5. Real-time monitoring 

The impact of network performance monitoring is widely recognized, with enterprises monitoring a long list of vital network metrics that give them visibility into network performance. However, it’s not always effective for every network. Applications on remote sites, such as oil rigs, and within large, distributed networks can often suffer from high latency that undermines the efficacy of the monitoring and causes alerts and insights to arrive too late for proactive interventions. 

With growing concern for security as well as performance for vital infrastructure, services, and applications on remote sites, we can expect to see increased support for an adoption of advanced tools that enable real time monitoring even in these challenging situations. AI monitoring can reduce and often eliminate latency for remote sites, while still centralizing all the data and insights through a single control admin dashboard. 

6. Cloud monitoring

Networks are increasingly moving to the cloud, if they aren’t there already, and even more are using hybrid networks which tend to have poor visibility. One major drawback to multi-cloud environments is that very often, each cloud system has multiple providers. 

This can require using a number of different Network Performance Monitoring tools, which can lead to even more complexity, and frequently results in more confusion than clarity. In response, we’re seeing the evolution of cloud network monitoring to become more centralized and vendor-agnostic, with support for hybrid networking. Newer cloud monitoring tools can aggregate data from every network in a single interface and deliver cohesive insights that allow teams to optimize the overall environment without penalizing one network for the sake of another. 

7. User experience

Last but not least, network performance monitoring is responding to a shift in focus from preventing downtime to improving end user experience. Netops teams are realizing that fast downloads, real time lag-free video conferencing, and low-latency connections are must-haves for enterprise network users, and they are seeking the tools to help them deliver. 

One result is the rise of using digital experience monitoring (DEM) to deliver visibility into communication paths, so that teams can identify and resolve issues that impact UX. Gartner® predicts that by 2026, at least 60% of I&O leaders will use DEM to measure application, services and endpoint performance from the user’s perspective, up from less than 20% in 2021.

We can expect to see this drive the adoption of solutions like synthetic monitoring, and web application monitoring, and greater concern for performance metrics like latency, jitter, throughput, and packet loss. At the same time, AI networking adoption is likely to rise, with Gartner estimating that it could improve network availability by up to 25% and enhance application performance to enable a better end-user experience.

Network performance monitoring isn’t standing still

It’s clear that network monitoring is evolving quickly, with performance becoming a greater concern and AI and automation bringing new capabilities all the time. Automation, cloud monitoring, user experience, security, and real time visibility are all poised to play key roles in driving next gen network monitoring software, which will result in greater security, more resilient networks, and happier customers. 

Download the Ultimate Network and Operations Monitoring Checklist 2023!