In an age where time demands are so high, "slow" has become the new "down" for both consumers and enterprises using web-scale content and cloud-based applications or services.
The solution to the challenge of protecting the web-scale content user experience lies in utilizing emerging analytics and automation technologies, which will bring carriers and web-scale content providers the agility, responsiveness and effective operational processes needed to support cloud and Internet of Things (IoT) applications and avoid slowdowns for end users.
Intelligent network services
One of the key ways to achieve this is through intelligent network services with real-time visibility, analytics and automation. These services will deliver superior control and optimization of both network resources and application traffic flows. An insight-driven, automated networking approach will also improve the user experience and enhance network security automation to protect, for instance, against potentially devastating DDoS attacks.
Globally, all web-scale content and IT service provider applications running over the Internet carry their content to end users over a complex network connectivity path. It includes multiple network domains operated by different companies -- from Internet service providers (ISPs) and content delivery networks (CDNs) to transit providers.
When users experience quality issues with these applications, it is a difficult, complicated problem to solve. From the data center to the end user, the application traffic may be impacted by network congestion at any point over multiple network domains. If there is congestion anywhere along the end-to-end, cross-domain network path, the end-user experience for specific applications may deteriorate significantly. These are the type of application issues that cause customers to switch ISPs or switch to an alternate application where performance is perceived to be better.
Leveraging analytics and automation for application performance
To network service providers, this may sound like a problem that is outside of the scope or control of any individual domain service. However, application performance isn't totally at the mercy of a "best-effort" Internet. Technology advancements in multi-dimensional analytics and insight-driven network automation deliver improved visibility and additional analytics for SDN automation and control. These technologies are not only economical, but also easy for any company involved to deploy without dynamically re-engineering their network.
Whether it's a web-scale content or IT service provider, an ISP, CDN or transit provider that needs to solve this issue, each provider can approach this challenge by correlating cloud and network data in real time to gain an understanding of how cloud applications and content are flowing to and through networks and more broadly the Internet. This data can then be leveraged to dynamically drive the underlying network control and optimization needed to meet fluctuating demand patterns. It can also be used to automate network service operations from provisioning and optimization, to assurance.
Sterling Perrin, principal analyst at Heavy Reading, published a related article in April that detailed some promising SDN use cases, and the top use case Sterling listed is network visibility and analytics. Sterling wrote, "Many operators are interested in 'early win' applications that understand both layers in the network in read-only fashion and are able to automatically correlate them into a unified view. This correlated view then can be fed into existing management systems to improve planning accuracy, identify anomalies in structure and make more confident planning decisions." (See The Value of SDN-Based IP & Optical Integration.)
Continuous monitoring of network utilization and performance needs to be tied to visibility of the multi-domain network topology -- including for network state and resource health -- in software-defined control and management systems. Only with this combined visibility are network operations able to achieve automation and assurance at cloud-scale and adapt quickly enough to rapidly changing network demand patterns. Network monitoring and analytics need to detect emerging issues, such as network congestion in real time, as well as have a cross-domain view of all network paths in order to optimally re-engineer paths or re-direct traffic flows dynamically.
This approach to network monitoring also involves harnessing a massive amount of streaming telemetry from network devices and servers, which must be combined with customer-provided data, application context and an understanding of the cross-domain and cross-provider paths that applications take through the Internet. The data must be correlated every few minutes to keep it fresh and relevant. By feeding these insights into carrier SDN and network assurance solutions, the result is a more intelligent software-defined automation for network optimization and application flow control.
Multi-dimensional analytics and insight-driven network automation is a new approach in the industry that not only meets the need for cloud speed, but also adds application-level visibility and control in the most cost-effective way. This approach will ultimately enhance the overall network value by protecting the end-user experience in real time.
— Sasa Nijemcevic is VP & GM of the Network and Service Management business unit at Nokia, and has 20 years of experience in the network management domain.