NFV and SDN are past their hype cycle and way into gaining significant traction since their initial introductions into the market. Many of our customers who have deployed or plan to deploy our solutions based on these technologies understand that an important factor for realizing the full benefits of NFV and SDN is service assurance. Service assurance is crucial in the era of digital transformation and in effectively operationalizing NFV and SDN enabled networks, towards offering your customers an outstanding, end-to-end digital experience. However, in today’s virtualizing network, service assurance requires a whole new approach and one that greatly differs from what traditional systems can provide.
First, let’s take a look at some of the important characteristics of the changing cloud infrastructure on which services are offered and what this means for assurance systems.
- Complexity and scale: Cloud infrastructures are extremely complex due to the large number of distributed network elements across multiple technologies and service domains. The multiple layers of the infrastructure stack which include the underlying physical elements, the virtual overlays, and service components creates critical inter relationships and dependencies. In order to rapidly isolate problems and proactively resolve them, first, the assurance system must holistically understand these links and dependencies and second, it must be scalable given the significant number of services it must support
- On-demand services and elasticity: NFV and SDN help deliver services agility and enable more flexible service consumption models such as the “pay as you go” option which is increasingly adopted by customers whose service requirements change over time. In this model however, the time between service turn-up and turn-down is much shorter compared to that of traditional services. This means the assurance system must be aware of the dynamic changes to service topologies, in real-time, in order to proactively assure services. Additionally, in NFV based infrastructures, service placement can also dynamically change, such as in failure scenarios. In such instances, the assurance system must again, quickly recognize and adapt to the associated changes.
- Multi-tenant environment: More and more businesses are turning to the cloud to experience new applications, collaborate on mission critical projects and utilize IT resources more efficiently. Here the service elements (compute, network, storage) that support these functions are virtualized on a per customer and per tenant basis. Therefore, it is essential for the assurance system to have an accurate and contextualized knowledge base on a per customer/tenant perspective in order to understand the potential impact a network issue may have. This is especially important for supporting those customers who rely on your services to run their most critical business functions
With traditional assurance systems, which are largely disparate management entities with operational boundaries and limited information exchange, addressing real-time requirements or deep service/tenant awareness is nearly impossible. This is especially true for high-density cloud environments where the assurance system must also effectively scale. Fortunately there is a new approach to assurance management Service Providers can adopt. This includes:
- Implementing an open, modular and extensible architecture for addressing cloud complexity: First, an open and modular framework will allow data to be accessed from a wide array of domain specific sources such as probes, EMSs and NMSs including those that support the existing physical domain. The purpose is to help you maximize the usage of what you’ve already invested in, while also allowing you to extend your capabilities into the new virtual domain.
- Utilizing big data platform technologies to enable operational scale and unified visibility: The big data approach helps mitigate the issue of operational siloes by aggregating and enabling analysis of data from various sources (both physical and virtual) that are specific to a service. In turn, it can paint a picture of the overall health of that service. Through the open framework, new and innovative analytical applications can also be added as you expand into new services and/or technologies.
- Integrating with service model driven orchestration system for real-time assurance: If the orchestration system maintains the latest device and service configuration information, it can integrate with the assurance system to extend the service model definitions to include the assurance requirements such as KPIs and other parameters relevant to SLAs. The goal here is to assure the service from the moment of instantiation. Even if there are changes to the underlying physical elements used to actually fulfill the service, the assurance system can dynamically adapt to these changes since it uses the latest information maintained by the orchestrator.
For more insight on the recommended approach to next generation service assurance, I encourage you to read this white paper, “The Role of Service Assurance in the Virtualizing Network” by Heavy Reading. Cisco’s goal is to help make your transition to NFV and SDN easier through our flexible, open, and modular offerings.
Please feel free to contact your local Cisco representative for more information and follow us on Twitter @CiscoSP360.