Delivering enterprise agility – how the CE services market will top $50bn in 2015
Date: Thu, 07/31/2014 - 13:06 Source: By Kevin Vachon, COO of the MEF
Increased enterprise agility and a dramatic boost to user productivity are promised by the latest wave of dynamic CE services now coming to market – a market set to grow by several percentage points in size to over $50bn globally in the next five years
Kevin Vachon, COO of the MEF
Image credited to NetEvents
Research and analysis firms Frost & Sullivan, Vertical Systems and Infonetics are all predicting an Ethernet services market worth approximately $50 billion by 2015, several percentage points ahead of the present market position.
The MEF (Metro Ethernet Forum) is the catalyst behind today’s $45B global Carrier Ethernet services market. At the MEF’s inception in 2001, the “metro Ethernet” market was fragmented into a number of services – e.g. Optical Ethernet, Switched Ethernet, and Metro Ethernet – with vastly different capabilities, often without carrier class capabilities or service level agreements (SLAs) and limited to “best effort” performance. The MEF created a collaborative environment, including service providers and network solution providers, to jointly define and standardize “Carrier Ethernet” towards today’s high quality service. By creating technical specifications, implementation agreements and certifying services, equipment and people, the MEF has enabled a holistic ecosystem responsible for Carrier Ethernet’s subsequent market growth.
What now, and what is needed
Packet-centric applications now dominate circuit-based applications and voice, video and data all share a common network infrastructure with the risk of conflict and service degradation. Voice communications is decoupled from the underlying infrastructure of telephones and PSTNs and runs as an “app” on devices connected to the Internet. No longer is the service either up or down, with the presence or absence of a dial tone: VoIP can suffer impairments such as echo or voice distortion, through dropped or delayed delivery of voice packets. A better service can be assured using private networks, but at the cost of reduced flexibility in terms of activation times and purchase models where the service providers require long-term leases to commit to the service assurances required.
We are, however, moving rapidly towards an even more dynamically connected future. Machine-to-machine (M2M) communications will push connectivity way beyond the number of connected humans, with connected cars, smartwatches and devices, tablets, intelligent control systems and sensors coming on line and communicating to automate our lives. Each of these applications will demand their own service levels, and degradation will be unacceptable in many cases. This will only be possible if the network infrastructure transforms to enable cloud and mobile services that connect people and machines in real-time, on-demand, with assured QoS and quality of experience (QoE).
As a practical example, consider mobile workers connecting over the Internet via IP VPN to their office network. This is fine for checking e-mail, swapping documents etc, but critical communications such as a videoconference can suffer degradation from other users sharing access or from congestion in the ISP’s network. It should be possible to request (and be billed for) a higher performance connection to the office just for the duration of the connection.
For the second example, an enterprise subscriber wants a network service to interconnect locations to their virtual machines (VMs) or Virtual Network Functions (VNFs) in a remote data center. This is only possible by using a number of transit service provider networks between the data center and the locations. So each of these network operators needs to orchestrate the setup of an appropriate internal networks and each of these operator-specific orchestrations need to be reconciled together to ensure the full end-to-end service required. Orchestration between the service provider and the cloud provider is required to automate the service ordering, provisioning, and management (OAM) of the virtual connections across each respective network and to setup the physical and virtual endpoints. This is a complex job that can take months, but should be delivered promptly on demand to meet real business needs.
To support agile business we need connectivity between physical or virtual endpoints with dynamic attributes to suit on-demand cloud services. Real-time applications that monitor performance should evolve to automatically request, or prompt the user to request, different classes of service as needed – eg reduced packet loss for the duration of a videoconference. The customer need only input basic information to order the service – e.g. service endpoint locations and service bandwidth in addition to billing information – in a manner similar to ordering cloud services, where components are ordered, fixed and recurring costs totalled, then the order is submitted.
Progress to date
The challenge of deploying networks across third party access vendors is already being addressed by a combination of existing technologies – Carrier Ethernet’s ubiquity and standardized connectivity; Software-defined networking (SDN); Network Functions Virtualization (NFV) and real time Big Data analytics to correlate data from the many network elements and OSS, and continuously analyse it.
Whereas barely 20% of Carrier Ethernet services succeed first time, and it can take over a hundred days to turn up a circuit, these principles have improved inventory integrity to over 90% accuracy, significantly reducing fall-out and improved time to market – while on-going automation of auditing and inventory updates is cutting OpEx. The solution began with a data audit extracted and mapped to a structured format – including OSS sources, activation notices, SLA agreements with AVs, excel spreadsheets and inter-carrier agreements. Automated continuous audit could now identify bad data and even assign a quality indicator to simplify integrity assessment. Continuous correlation plus big data analytics identify risky changes, and check consistency and value ranges, and warnings are transmitted to the data owner. The system also provides a graphical overview of the topology, revealing actual circuit inventory details, simplifying ordering, provisioning and service assurance.
In a second example, workflow automation is cutting costs and accelerating service turn-up, leading to rapid growth of the provider’s footprint and capacity. An Additional Services Request (ASR) – eg Move, Add, Change or Delete network functions – is transmitted to the access vendor by web form, and changes are automatically broadcast to all network elements, without delay or risk of human error. This includes populating test equipment with updated test configurations so SOAM tests run automatically, and results are collated and reported.
In a third example real-time feeds are taken from existing monitors and summarized in a single customizable dashboard – registering alarms and correlating them to circuit segment states. Thresholds are set for each access vendor and used to benchmark SLA performance so reports can indicate exception events and leverage historical data to determine trends. Without manual work, the provider now benefits from lower MTTR, faster triage and root cause analysis – thanks to rapid, accurate isolation of degradation and better SLA penalty capture with authoritative proof and reporting
The Carrier Ethernet market has reached a turning point. The victim of its own success, it has given business a taste of global networking benefits, and is now struggling to deliver those advantages as seamlessly and fast as agile business requires.
The MEF is aware of the need and the challenges, and is laying the framework to enable new types of network connectivity, better aligned with cloud services and opening up new revenue opportunities for service providers and the ecosystem of network solution providers. This is good news for the enterprise and, ultimately, for the global economy.