Network Visibility & Data Veracity – the Keys to Operational Simplicity
James Crawshaw, Senior Analyst – Service Provider IT and Automation, Heavy Reading
The Seven Bridges of Königsberg was a mathematical challenge that the intelligentsia debated in 18th century Prussia and beyond. Swiss mathematician Leonard Euler proved there was no solution to the problem (a route through the city that would cross each bridge exactly once) and in doing so laid the foundations for graph theory. Two hundred and eighty two years later, only five of the bridges remain in Konigsberg (now known as Kaliningrad) and it is now quite simple to show that a Eulerian path is possible. However, if we were looking at the old map we would not realize this.
The situation is analogous to the problem facing service providers today when their inventory management systems no longer reflect the reality of what physical infrastructure is deployed in the field (devices, cards, ports) and what services are currently enabled on that infrastructure. This not only causes problems with service configuration and implementation but also with diagnosing and troubleshooting problems in the network. It is simply too hard to find the right information at the right time. Operations staff and customer service teams often end up bouncing around from system to system in search of the answer.
Many operators are in the process of adopting the SDN and NFV paradigms in order to make their networks more easily managed and automated. However, they have not necessarily implemented the requisite tools to automate operations. There is little point speeding up provisioning times via new orchestration systems if processes such as fault or incident management, problem detection, troubleshooting, service testing, service assurance and change impact analysis remain slow, manual and reactive.
Operators are also exploring how machine learning algorithms can be used to increase the efficiency of network operations. However, if the data with which these algorithms are trained is inaccurate, these initiatives are unlikely to deliver much value. To solve these problems, operators need tools to improve network visibility and data accuracy. Such tools should bring visibility, veracity and simplicity to the customer care, network operations center and planning teams.
The visibility aspect involves an end-to-end topology view of both physical and logical connectivity that spans all network domains (access, metropolitan, transmission, core) and all suppliers within those domains. The veracity component is about managing and maintaining data integrity and quality. Most operators suffer today from incomplete network inventory systems that diverge more and more from reality as time passes and changes occur but are not recorded. What is needed instead is a consolidated blueprint of the network drawn from order management systems, network inventory systems, element management systems and network management systems.
The simplicity facet relates to simplified operations. The key to this is providing a single source of inventory "truth" regarding network elements, cards, ports and services. This avoids the "bouncing around" between systems referred to earlier, speeding up the time to resolve problems.
Like the Konigsberg Bridge problem, operators need an accurate, up-to-date network and service topology view in order to determine the "Eulerian path" of service configuration and implementation. This topological view or model must be across the various network domains (access, core, etc.), across vendors, across legacy as well as next-generation networks and services, and must be reasonably close to real time.
Real-time topology enriches and contextualizes the mountains of network data ingested (faults, test data and logs) in relation to services and customers, facilitating service impact analysis, common cause analysis and change impact analysis. It provides customer-to-infrastructure dependency analysis, making service assurance more proactive. For example, we might be able to predict how network changes will impact a particular customer and, as a result, take pre-emptive measures to ensure continuity of service quality.
A better understanding of the end-to-end infrastructure will enable CSPs to provide their customers with an improved quality of service and support, leading to lower customer churn and, potentially, market share gains. It will also eliminate the time wasted investigating discrepancies in data held in different network inventory and management systems. This will increase the efficiency of customer support, planning and provisioning processes, allowing operators to reduce operating expense and more fully embrace automation.
This blog is sponsored by EXFO (Nasdaq: EXFO; Toronto: EXF).
— James Crawshaw, Senior Analyst, Heavy Reading