Network automation is both a process and a journey and its state within the enterprise continues to evolve over time. Organizations have historically managed their network from a human-centric worldview, which revolves around CLI and Graphical User Interfaces. This manual approach results in a high tax on IT teams due to the human-heavy requirements of managing the network infrastructure and low-quality data due to human augmentation and errors. Today, organizations are moving towards a machine-centric and automated worldview of the network, which is first and foremost programmable. In a modern network, data is highly decentralized and federated. This machine-centric approach embraces model driven behavior and an open API ecosystem at its core which allows an easier pathway to network management as enterprises move towards a more connected and cloud-centric view of network services.
A recent study conducted by Enterprise Management Associates (EMA) reports that over 60% of enterprise organizations today still utilize ad-hoc manual and reactive script-based solutions for network automation. Unfortunately, the general market perception is that there are only two approaches to network automation, human-centric and machine-centric, and that they are mutually exclusive. Either the NetOps teams continue with ad-hoc scripts or they have to embark on a painful rip-and-replace initiative to leverage the programmable network through a vendor supplied tool. Such a mutually exclusive approach is not practical or feasible for most enterprises.
As network operations and management continue to evolve, there must be a transitional pathway that allows organizations to leverage existing CLI network automation efforts, while still making progress towards true closed loop automation solutions without having to start from scratch and re-work all previously built automations. While it is becoming clear that adoption of controllers and orchestrators represents the future of how networks will be managed, organizations still need to utilize their existing technology investments and intellectual capital as their automation journeys evolve and progress.
Integrated Approach to Network Automation
As an organization considers its automation journey and an integrated strategy required to evolve, there are four key recommendations as the course is plotted:
Leverage Existing Automation Efforts
Identify existing CLI automation efforts and integrate these scripts into modern network architectures by using available technologies that transform scripts into APIs. One example technology is the Itential Automation Gateway.
Expand Automation Participation
Enable network engineers, software developers and IT operations to participate in network automation through low-code, drag-and-drop environments. The more groups that participate in network automation, the faster the evolution will progress (permeance of the technology) and be embraced (company buy-in) while achieving cost benefits and risk reduction (reduced human error).
Scale Automation as You Grow
An important part of the network automation evolution or journey within an organization is to deploy at a pace that best fits the operational and business goals. Whether it is basic operations and maintenance tasks such as router/switch upgrades or more advanced network automations such as managing device configuration and compliance, service management and policy management. Taking a pragmatic approach and understanding the pace at which the organization will best handle the transition towards automation is critical.
Integrate with Network Ecosystem
Extend automations beyond the network by leveraging APIs in order to connect other tools within the IT ecosystems of the organization. Network automation in a modern world is not limited to changes required on network elements, but across several ITSM systems that are responsible for activities such as sales orders, change management, inventory, ticketing and notifications. It is essential for organizations to have a solid strategy around achieving end-to-end network automation in order to realize the ultimate benefits of an automation strategy.
As each organization considers its unique journey and transition towards network automation, it is important to understand that it should not require a disruptive rip-and-replace approach in order to achieve the benefits of automation. A better venue would be to complement the human-centric (past) and machine-centric (present) approaches to network automation as organizations evolve and transition into the future. Ultimately, this integrated strategy that utilizes existing investments in automation (CLI) alongside programmable, decentralized, federated and model-driven technologies will enable more groups to participate in network automation, lowering operational costs while mitigating risk.
Get a more in depth look at Itential’s ‘Bring Your Own Automation’ capabilities in this on-demand webinar to see how your organization can make the move toward a modern approach to network automation using your existing infrastructure and investments.
Karan Munalingal
Head of Solutions Engineering ‐ Itential
Karan Munalingal is the Head of Solutions Engineering at Itential. Previously, Karan ran systems engineering at Ciena, focusing on carrier ethernet and core switching platforms. At Itential, Karan drives automation strategy serving global customers transitioning to modern networks and software-defined network programmability.