David Mihelcic is the federal chief technology and strategy officer for Juniper Networks.
When you think about the network of the future, what do you envision? If you’re just thinking “automation,” you may not be thinking big enough. That’s because, as with most things in technology, today’s innovation in automation is really only a stepping point to what comes next. For federal IT, “next” means intent-based networking.
Intent-based networking uses high-level business outcomes to drive network configurations and performance. It’s much different than the traditional approach involving granular, difficult to understand, machine-level specifications. An intent-based network allows teams to determine and define business outcomes, which are then compiled into detailed configurations that are automatically implemented by the network.
» Get the best federal technology news and ideas delivered right to your inbox. Sign up here.
In the very near future, intent-based networking will allow government agencies to achieve better business outcomes with minimal manual input. It can speed up operations and service delivery, reduce network fragility, and potentially help save on operational expenses. Remediation time will be reduced, as networks will quickly and automatically identify and resolve faults in minutes, not hours or days. Cybersecurity postures will improve significantly, as networks will be better able to discern hostile activity and immediately initiate actions to mitigate potential threats.
Intent-based networking will also open up doors for federal IT professionals. Individuals who have traditionally spent inordinate amounts of time “keeping the lights on” will be able to focus more on the creation of initiatives designed to bring true value and innovation to their agencies.
There are several critical technology factors to consider when laying the groundwork for intent-based networks. Let’s take a closer look at each of these and the roles they will play in developing your network of the future.
You must possess a deep understanding of everything that is going on in your network on a global basis across all network elements and their supporting systems. This understanding is critical to meeting business intent. Therefore, intent-based networks will require active telemetry and anomaly detection based on machine learning. This will be used to detect changes in network conditions and drive remediation of faults before they become customer impacting.
When those changes occur, automation will proactively configure adjustments to ensure the network continues to meet overall business intent. Network responses will become inductive and based on machine learning, greatly reducing (or even eliminating) the need for human guesswork and interaction. Service placement and motions will become automatic, and upgrades will be automated based on configured services.
It will no longer be enough to rely on specific configuration parameters such as destination IP addresses, ports and protocol types. Instead, you will need to be able to assign priority to applications. If there is contention for resources, the network will be able to understand which applications receive priority. This will enable the network to automatically allocate resources and provide the most cohesive user experience.
Autonomous decisions will need to be made in response to changes in the original business intent. Teams will need to have the ability to update their global network configurations automatically and in real-time in order to get things back on track. Therefore, today’s rigid, rule-based systems involving simple manual programming commands will evolve into machine learning. Decision-making will advance from static programming to algorithms that learn from data inputs. Networks will be able to make predictions and decisions and take appropriate actions.
Local and Global Views
Instead of just asking, “How do I fix the packet loss on this router?,” intent-based networking managers will be asking, “On a global basis, is this network meeting my agency’s overall business intent?” Answering this question will require teams to collect metrics and feedback data on both a local and global network basis. They will need complete views into the performance not just of individual network elements, but the entire system.
SDN as a Foundational Element
Software-defined networking, known as SDN, is a core foundational element upon which government agencies can build toward their intent-based future. SDN helps centralize network management and control and is already an ideal solution for improving agility and facilitating network automation.
SDN with an added intent-based layer can take a declaration of intent and translate it into service-, device-, and technology-specific semantics. In effect, intent-based models focus on declaring “what” the network should do; SDN with an intent-based layer tells the network “how” to do it. That makes SDN a cornerstone for automated, highly efficient intent-based networks.
Vendors have already begun building towards this future. Today there are modern networking technologies that allow teams to specify high-level performance goals for wide-area networks. They take overall intent and continuously optimize the network to ensure it’s delivering the requisite level of performance.
In short, you can stop dreaming about your future network, and start building toward it now. The foundational elements are already in place to create an intent-based network that will deliver substantial cost and performance benefits.