Network monitoring is the art of analyzing an IT network’s status and performance for optimal reliability. The tools and techniques I use have evolved over time to best support today's vast ecosystems of modern hardware and virtualization technologies. Since the Internet’s inception, when the first Unix-based systems came together to form a network, there’s been a need to monitor networks. Tools and technologies have come a long way in the past 25 years that I've spent in this discipline.
Today's network monitoring tools gather massive amounts of data to help organizations better measure, diagnose, and optimize the quality of service and, ultimately, the user experience of their network. By assessing different types of network data (such as packet data, network flow data, metrics from network infrastructure devices, synthetic tests, and more), organizations get a clear picture of the reliability of the network—and what the experience is like for end-users.
The network landscape has also undergone dramatic change over the years. Monitoring has kept pace by continually evolving alongside these network advancements. Most notably, the size, scope, and complexity of networks have increased significantly, and it's now more important than ever for companies to ensure their networks run reliably.
Let’s take a look at the technologies that make this possible today, which will only continue to add value in the future.
1. The devil is in the data, but automation helps
In today’s world, organizations have the ability to use data to learn more than ever before. You can now capture, store, and process colossal amounts of data with the help of automation. By its nature, network monitoring captures massive amounts of data, and the definition of "massive" is always growing. This creates the difficult and costly problem of developing infrastructure that can keep up to support big data.
Only now, because of advancements in big data tech and automation, it's possible to look at all of the data. We can now gain insight into the behaviors of the network in ways that were previously unimaginable. For instance, having insight into all of the data offered by just two basic KPIs, we can quite reliably tell a company everything it needs to know about its network. Put simply, automation can generate insights—at speeds—that aren't otherwise possible.
As automation technology continues to improve, organizations will be able to collect, store, and analyze network data—at scale—leading to insights that we can't imagine today.
2. Security advancements
Network monitoring systems need access to critical infrastructure, making them common vectors for cyberattacks. Given the increasing number of security threats organizations are contending with today (we all remember the recent Solarwinds cyberattack), it's even more crucial to ensure networks are as secure as possible. Thankfully, the increased adoption of things like zero-trust security frameworks helps to further bolster network security. This approach requires all users—both in and outside the organization's network—to be authenticated, authorized, and continuously validated prior to being granted access to applications and data.
In the future, we can expect to see more organizations get on board by investing in zero-trust principles.
3. Open source contributions
As long as networks have needed to be monitored, open source contributions have propelled the space forward—both in terms of technology and the ability of the community to act collectively.
In terms of technology, there's no better example than SNMP, the community-built, standardized protocol for network management. Monitoring has relied on SNMP since the beginning, and we still haven't seen another standard emerge that has the support or capabilities that SNMP does. That said, we have seen recent open source projects like OpenConfig/gNMI and OpenTelemetry that approach monitoring in new and exciting ways. Ultimately, the technologies that make the most sense will see mass support and use—propelling them to become the next standards and frameworks. I think we'll see a lot more maturity and progress out of these projects, and I look forward to supporting them along their journey.
The other side of open source is the community aspect. Building network monitoring solutions requires the ability to test against all possible hardware, devices, and configurations—a lab, of sorts. The importance of open source is invaluable here. No single company could ever do this alone; the community is the lab. Having a community to run changes, create new code, and contribute feedback is critical. Without this, our ability to test things under the conditions of thousands of different deployments and scenarios or configurations wouldn't be possible.
Open source will absolutely be essential in the future of monitoring.
4. Developments in real-time monitoring
A long-time goal in network monitoring is to reduce or eliminate latency—the time between a device or system generating data and the monitoring tool receiving and digesting it. This will become more critical for infrastructure, services, and applications located in remote sites within large, distributed networks. (For example, consider an IoT sensor on a drilling platform in the middle of the ocean.)These can be hard, if not impossible, to reach and monitor from a central location such as a data center or the cloud. It's likely that the industry will continue to make technical improvements that enable monitoring mechanics and capabilities to be deployed almost anywhere while keeping the monitoring logic, configuration, and administration centralized for easier management.
So, as we look to the future of network monitoring—and monitoring in general, I believe that contributions from the open-source community, in addition to developments in automation, security, and real-time monitoring will come together to make network monitoring more reliable and efficient than ever before. We can expect to hear much more about these technologies—and how they're transforming network monitoring—in the coming years.
David Hustace is CEO of The OpenNMS Group, Inc., a subsidiary of NantHealth, Inc.