Automating network lifecycle management using AI

Managing devices and policies on a traditional network is a complex, manual task. So, what if we could automate the process leveraging AI?

Traditionally, networking has relied on manually setting up individual policies for network devices. Because each vendor’s interface and syntax were unique, network administrators and team members had to learn the syntax for each piece to configure the network.

Today, implementing business requirements still requires significant human interpretation and manual intervention to ensure that networks and IT systems meet business needs. In most cases the process is lengthy, resource intensive and error prone.

With such complexities, repairing an outage, making routine changes, delivering new services and meeting service level agreements are huge challenges. This is not what an agile digital business environment looks like.

Further, business users must also use technical language to define their network requirements, even for simple changes. For example, if a business user wants server X to be accessible from any server in the domain, he would need to specify that he wants connectivity to server X from servers A, B, C and D. 

Translating intent into policy

Now imagine if business users and network administrators could define their needs in simple English, and an AI-powered engine translates each need, or intent, into a network policy, implements the policy and automatically monitors continuously for assurance.

DXC Labs is working on such a proof-of-concept, which is designed to automate network lifecycle management. The figure below shows how networking shifts from manual and serial to automated and continuous.

Comparison of current networking and future networking

Click image to enlarge.

Continuous monitoring and remediation

Key to success is the continuous cycle of verification and remediation that constantly checks if the configuration meets the intent and makes corrections in real time.

Our system gives the user a single interface to seamlessly manage all network devices. It allows machines to take care of machines, freeing network administrators to design the organization’s network policy. Such a system automatically enforces network policies and ensures a stable and secure network.

Users can describe their intents using spoken language. For example, a user could say, “I need maximum bandwidth available on the payroll servers for the second half of every month.”

The system can also receive intents (requests) directly from service management tools, leveraging the API for integration with those tools.

The system’s components address the following functions:

  • Business intent: Provides a list of intents in a catalogue the user can choose from, and enables the network team to monitor the operations and intent requests on a dashboard in real time.
  • Translate: Translates the “what” into the “how.” The interface is a menu-driven graphical interface or an interface that uses natural language (currently English in our proof of concept).
  • Collaborate: Provides policy recommendations, along with tradeoffs, for every intent. These recommendations are powered by AI and machine learning. Users choose the best recommendation based on their needs.
  • Activate: Deploys the policies throughout the network by automating systemwide changes to all relevant network and security devices.
  • Assurance: Reacts in real time to changes, such as a link failing or a device going offline, by continuously gathering data and monitoring the state of the network. Uses machine learning to identify the best way to maintain the desired network state and take automated corrective action to maintain the network state. Live vulnerability feeds, ingested as intents, keep security policies up to date.

A typical workflow might look something like this:

Workflow for automated network management

Caption: This workflow for automated network management starts with business intent and applies AI to convert the intent into a network policy, monitor the network and maintain its desired state. Click image to enlarge.

In our next blog post, we will discuss how this proof of concept can be applied to specific business scenarios.


ArunKumar Amarnath is a technologist in DXC Labs India who focuses on intent-based networking. Based in Bengaluru, his experience includes solution development, Agile, DevOps consulting and system implementation. @arun_amarnath

 

 

Baskar Venugopalan, associate director at DXC Labs India, focuses on seamless and secure integration of business, information technology and operational technology. Based in Bengaluru, his three decades of experience include working with cloud and big data, leading incubators, and partnering with multinationals like Microsoft and Oracle. @venugbx

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: