For many government agencies, the road toward network modernization has been filled with more twists and turns than the racetrack at the Indianapolis Motor Speedway. From data center consolidation to making due with less, federal IT professionals have had to balance the government’s desire for greater efficiency with the need for speed, resiliency and security.
Self-driving networks can help agencies balance all of these goals. They combine telemetry, workflow automation, development operations and machine learning to create a more responsive, adaptive and predictive infrastructure.
True to their name, self-driving networks are highly autonomous and function with limited human intervention. They can self-analyze, self-discover, self-configure and self-correct. They can instantly respond and adjust to changes in network activity — a possible cybersecurity threat, for example — with little need for manual intervention.
In addition to improved security, self-driving networks can also reduce operational expenses, increase network reliability and empower federal IT staff. An automated network greatly reduces the costs associated with the time and resources it takes to manage network functionality.
Self-correction allows the network to continue to operate at maximum efficiency without manual oversight. Employees can spend less time attempting to fix network problems and more time writing the code that will help move their networks (and their agencies) toward the future.
We are continuing our journey toward self-driving networks. Before we reach the end, we will have to make a few stops along the way.
Mile 1: Human-Driven Automation Gets Agencies Started
Some agencies are already moving away from manual legacy systems in favor of modern networking technologies that require less hands-on maintenance. As part of this effort, teams have begun leveraging common scripting languages to automate daily tasks.
This is the beginning for self-driving networks in federal IT. In taking small steps toward automation today, managers are paving the path forward toward the self-driving networks of tomorrow.
Mile 2: Event-Driven Automation Sets Network Rules
Human-driven automation will ultimately lead to event-driven automation. At this point, managers will be able to integrate data sources from telemetry and establish rule-based actions that will enable their networks to automatically respond to an event.
There will still be some element of manual operation required during this stage, but the network itself will do most of the work. IT managers will define the rules, but the network will act to enforce them.
Mile 3: Machine-Driven Automation Drives Intent-Based Networks
In this part of the process, managers will be able to develop algorithms that drive the behavior of their networks. They can begin to move away from designing specific configurations and toward creating intent-based networks.
They will no longer have to tell the network how to achieve a particular outcome; instead, they will simply need to declare specific intents — network goals or objectives, for example. The network will self-configure itself to achieve those objectives.
Mile 4: Achieving the Self-Driving Network
As we reach our final destination, we will begin to see almost completely automated networks based on augmented and artificial intelligence. These networks will be able to automatically adapt to events and conditions within and around themselves with minimal human intervention.
Software-defined networking (SDN) can be the catalyst to get started by supplying a means to automate network management tasks and processes.
Federal IT professionals can even take things a step further by implementing SDN with an intent-based networking layer, which will allow them to take a declaration of intent and translate it into service, device and technology-specific semantics. It’s not a fully self-driving or intent-based network, but it is a start toward a more efficient, secure and reliable network.
This content is made possible by FedTech. The editorial staff of Nextgov was not involved in its preparation.