, Regular Visitor
Network Services, at Your Fingertip
Oct 17, 2013


iPhone is no doubt one of the most successful products in the history of technology. Aside from other factors, the App Store provided a key differentiator that made it hard for other smartphone vendors to replicate. In July 2009, within a year of its launch, 1.5 billion applications (apps) for the Apple iPhone had been sold from its App Store. This was considered quite an achievement considering the fact that Apple's highly popular 'iTunes Music Store' had only managed to sell about 70 million songs in its first year, and had reached sales of a billion only after three years.


Ultimately, the App Store provided a platform for new innovations in mobile software application. Beyond the smartphone hardware itself, the broad selection of apps is a major reason why consumers want the iPhone, and the sticky factor for consumers to keep buying iPhones.


Can the iPhone App Store success be replicated by telecom service providers? With a little bit of help from modern networking technologies, it is possible!


One might ask, what are the challenges that telecom service providers are facing?


Service providers are challenged to monetize their networks and lower operational costs under competitive pressures from OTT providers. They need to provision new services to reach new revenues in a very agile manner over their existing infrastructures.  They also need analytics to gain insights into the services and tune down or totally repurpose the resource associated with unsuccessful services.


But telecoms networks contain an increasing variety of special-purpose hardware appliances. To launch a new network service often requires yet another appliance and finding the space and power to accommodate these boxes is becoming increasingly difficult, in addition to the complexity of integrating and deploying these appliances in a network. Moreover, hardware-based appliances rapidly reach end of life: hardware lifecycles are becoming shorter as innovation accelerates, reducing the return on investment of deploying new services and constraining innovation in an increasingly network-centric world.


Network Function Virtualization (NFV) evolves standard IT virtualization technology to consolidate many network equipment types onto industry-standard high-volume servers, switches and storage. It involves implementing network functions in software that can run on compute hardware, and that can be moved to or instantiated in various locations in the network as required without the need to install new equipment.

NFV brings significant benefits for network operators and their customers including:

  • The ability to provision and scale out services quickly and cost-effectively on existing infrastructure, enabling operators to better compete with Over-the-Top offerings.
  • Reduced time to market and improved return on investment on new services
  • Reduced operator CAPEX and OPEX through reduced equipment costs and power consumption.
  • Openness to the virtual appliance market and pure software vendors for new innovations


NFV is highly complementary to Software Defined Networking, aka, SDN, once network functions are virtualized, they can be dynamically instantiated and chained in the required sequence by the SDN controller.


NFV and SDN set the foundation for service providers to build a network services platform to foster innovations of new services from a much wider range of vendors. The end result? Service providers can launch “Service Stores” for their customers to shop for new services and order services online.


So what happens behind the scene once the customers finishes the ordering process online? Let’s use Juniper’s Contrail as an example to show how services are provisioned based on customer requests.


Contrail exposes RESTful northbound APIs to cloud orchestration platforms or OSS/BSS platforms to receive service creation commands. It then spins up one or multiple virtual machines on standard servers running software images of relevant network functions.


Contrail vRouters run along with the hypervisors on these servers and communicate with Contrail controller to dynamically set up overlay tunnels to route traffic between the designated networks through the newly created service.


When demand for the service increases Contrail can scale out the service to more virtual machine instances and load balance among them. Similarly when the service demand comes down, Contrail can scale back on the number of service instances to conserve resources, achieving truly agile and elastic service provisioning.


This almost sounds too good to be true, and one can’t stop wondering when this will become true. The answer is NOW. Contrail is production-ready and you can see this NFV and service store demo yourself.