Operators have a number of requirements to help ensure the smooth operation of their networks. These include access to end-to-end visibility of service performance and the underlying network that supports it, and also the ability to correlate between service impacts and resulting network failures. Without this knowledge, troubleshooting will be too slow and too reactive to avoid more failures. In turn, this will lead to growing disappointment for customers (and managers) in the new revenue generating services, such as on-demand networks that are just beginning to emerge.
Of course, networks can never be perfect, as to be perfect would require the perfect budget — budget constraints must necessarily have an impact. This means that it's quite a challenge to achieve the end-to-end visibility required for service assurance.
Real networks do not provide all the information and control engineers and operations teams need to support rapid troubleshooting. Not all sites, interfaces, elements and subscriber databases are covered by typical instrumentation built on passively collected measurement data, resulting in dark areas and further challenges relating to control and user planes.
Advancements in passive service assurance platforms are improving their ability to correlate and drill-down on the data they have access to. However, that data is massive in volume and velocity, and at the same time incomplete, which makes finding the source of service degradation much more difficult in practice than in an idealized architecture. NFV introduces new challenges, with dynamic scaling and with the lack of standardized interfaces for data access. For example, logical interfaces that use communication between functions, hosted on the same server, will not produce signaling in the physical interface, a new issue for systems designed to capture that signaling.
So what is the solution to imperfect networks and non-standard interfaces? The answer is to combine passive data collection systems with a service assurance platform that utilizes active monitoring.
To learn more about network impact due to virtualization or VoLTE implementation, watch our recent webinar - "Can you identify the causes of your IMS network outages?".