Cisco Blogs
Share
tweet

Prognosticating a Cloud Native Mobile Service Provider

- March 17, 2017 - 0 Comments

Written by Vy Kuchibhatla, Senior Manager, MBG Operations and Strategy

At MWC 2017, two key themes received a bulk of the focus from a network equipment provider perspective: 5G and NFV/Automation (Verizon/Cisco, Telus/Cisco, NTT/Cisco, Reliance Jio/Cisco). This article focuses on the genesis of NFV/Automation, its place in the Cloud and some key gaps that need filling.

To understand why NFV/Automation is so very key to Mobile Service Providers (MSP) everywhere, let’s dial back in time to 2014. Gartner had then predicted that by 2017, 50% of Global Enterprises will have implemented Web-Scale IT architecture. As proof positive, Cloud Native pioneers Amazon, Google, Netflix, Facebook have been and continue to implement enterprise capabilities at speeds and agility that belie their size and capacity, while at the same time reducing TCO and touting OpEx metrics that grabbed the attention of their Service Provider brethren, whose traditional networks, by comparison, require months of testing and pre-staging for feature and / or capacity roll outs.

So, what gives?

Although MSP networks adhere to standards from 3GPP, IETF and other bodies, traditional networks nodes are characterized by bespoke architectures, usually meant to reside on purpose built hardware. During the 2G, 3G, and even 4G, evolution, MSPs focused on network and capacity expansion, stoked by competitive TTM for service offerings. Sufficient ARPU margins fueled the expansion of highly custom mobile networks.

As the global wireless industry looks to overcome the commoditization of bottomless connectivity, MSP operational focus has shifted to resource optimization. Driven by revenue stagnation, the need for OTT use cases to drive top line will further increase customer-centric experiences. To address this dichotomy and in an effort to transition to efficient elastic networks, the concept of NFV/Automation has taken a stronghold as MSPs seek to adopt Web Scale-like DevOps deployment models to provide true Mobility-as-a-Service (MaaS) offerings.

However, SPs have in general struggled with providing competitive cloud based services. Unlike Cloud Native applications and simple VNFs, typically consisting of a single Virtual Machine (VM), SPs have taken “lift and shift” approaches to partially cloudify their networks or use cases.. Others are vying for private Cloud offerings as an overlay to hybrid networks with both VNF and traditional nodes. The resulting VNFs have multiple distinct VNF Components (VNFCs) with a variable number of VMs depending on feature optioning and desired performance specifications.

As traditional networking equipment vendors help MSPs scale to enterprise NFV/Automation networks, two key areas of focus emerge:

  1. Build for scale – religiously: Ease of instantiation and scaling is a core tenet of Cloud Native apps. By contrast, each mobile subscriber is a tenant unto themselves. Facebook, a $400B market cap company managing operations with only 17K employees, religiously adheres to design replication and elimination of one-off architectures, driving $23.5M of company value per headcount. Traditional “network function” (note the shift in terminology from “network equipment”) vendors need to move to a Cloud Native re-design to enable full realization of TCO benefits through Lego block like micro-services and service chaining.
  2. Automation focus: Ease of instantiation and scaling are core tenets of Cloud Native apps. ETSI MANO references the orchestration (NFVO) layer to drive service-level automation down the stack with network function descriptors driving config changes through the VIM and VNFM. Simplifying this management flow by embedding operational simplicity of VNF on-boarding, deployment and management is essential to deploy low cost multi-vendor VNF networks. This can be achieved through standards based VNF lifecycle management automation tools and is a necessity to ease the MSP transition to a fully Web-Scale like deployment.

Cisco’s Ultra Services Platform introduces Ultra Automation Services to address VNF-level automation from IT-prep, install, instantiation and deployment. The UAS roadmap extends this automation to instrument, validate and apply SLAs to VNF performance. Watch for a white paper detailing the value proposition and architecture of UAS.

Additional Resources:

 

Tags:

In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.

Share
tweet