Telecom IQ
Showing results for 
Search instead for 
Did you mean: 

NFV Nuts & Bolts: Multiple Operating Environments for Communication Applications


Author: Tariq Khan, Chief Technologist Cloud/SDN, NFV


The fundamental concept of Network Functions Virtualization introduces the paradigm of abstraction of the network functions from the underlying platforms that can help operators realize their goals of operational efficiencies, business transformation & financial benefits. This requires the operators to architect, design and operate the platform that will be used to host the virtualized network functions (VNFs). This blog looks into classifying the characteristics of different operating environments that are required to host both the current and future network applications.


Given the growing maturity & adoption of cloud capabilities and delivery models comes considerable debate about the “right evolutionary path and/or model” to design, develop, characterize, integrate, deploy, operate and assure network based applications.


While facing ever increasing competition from cloud service providers (i.e. Amazon, Google, etc.) operators & communication service providers have initiated efforts to re-establish themselves as a leader in offering a diverse and differentiated set of network based applications & services that generate considerable revenue. Furthermore, unlike a brand-new startup, operators do not have the luxury of rewriting core applications – i.e. a Greenfield – while not taking in to consideration the existing operational, financial and organizational impacts.


In the near term, it may be necessary to drive and maintain multiple operating environments, each of which is suited to different categories of applications and services, with a view to evolving towards a completely software-based cloud delivery model.


Different operating environments that need to be considered:

  1. Traditional Operating Environment – This is characterized by the existing applications/services that are delivered by Verizon’s incumbent vendors as physical network appliances – e.g. FW’s, mobile GW’s, load balancers, IMS core, etc.
  2. Cloud Native Operating Environment – An application that is specifically designed for a cloud development and operational environment - as opposed to simply being migrated or ported to common compute infrastructure with a vertically integrated SW stack (OS, Linux, Controllers/Managers, etc). These applications are highly agile and highly available, make use of common infrastructure components such as message bus/queue, shared databases, information datasets, etc.  The application itself is very “lightweight” and relies on the cloud platform to provide scaling, communication, performance, reliability and agility.
  3. Cloud Embracing Operating Environment – This Operating Environment is what is most familiarly referred to as “Private Cloud” in many Enterprises. It is characterized by homegrown and vendor-delivered applications.  An application that was developed using traditional architectures but has been migrated and validated to be deployed on cloud architectures. These applications typically do not have any dependency on underlying hardware
  4. Communications Cloud Operating Environment – This Cloud Operating Environment differs from classic enterprise clouds, in that it is intended to host real-time applications spanning multiple datacenters and wide-area networks. Given the business criticality, complexity and latency sensitivity of these applications, there needs to be considerably greater attention paid to topics such as network agility, service chaining, integration with OSS/BSS and change management than is typical of the Cloud-Embracing Operating Environment. This has huge implications for the availability and performance requirements of the underlying Virtual and Physical infrastructure layers as well as functional adjacency.


In the near term, it is expected that a single network service may have requirements such that its constituent VNFs are hosted in multiple operating environments. As an example, the VM’s hosting the control function of the network service may be hosted in the cloud embracing operating environment and the forwarding functions hosted in the communications cloud operating environment. An example would be an IPTV service – the access router that is common across services may be in #4 (communications cloud environment), the transcoder in #3 (cloud embracing environment) and a cloud native ad delivery engine in #2 (cloud native environment).


Each of these Operating Environments have different characteristics and impose different requirements on the underlying platform & infrastructure as shown in the table below:



The industry is moving towards the Cloud Native Operating Model because it provides for resiliency at the application level, and the maximum flexibility of hosting applications in private, hybrid or public clouds based on cost and other considerations.  This is the tallest order for the evolution of network based functions and should be the ultimate goal – understanding that a “hot cut” is not possible and that an evolution (how the products, services and functions are sold/consumed, developed, tested, deployed and supported) is necessary.


There is a general consensus that a move to cloud native operating model has the potential to provide the most long term benefits in terms of agility and costs, operators and their network function suppliers, iy may not be able provide applications that align to this architecture in the short term. However, operators needs to provide this as an option for early adaptors (both for internally and vendor developed applications) to facilitate unconstrained innovation especially for applications that leverage the SDN approaches.

Operators are likely to realize the above three operating environments from OpenStack as following:

  • Cloud Native operating environments can be deployed using the top-of-trunk OpenStack tree using all open source or subscription free flavors (or low cost flavors) on top of white box hardware
  • Cloud Embracing operating environments can be deployed using the mainstream enterprise OpenStack distributions using a distribution from a vendor that provides a supported offering on top of enterprise class hardware platforms.
  • Communications Cloud operating environments can be deployed using a premium version of OpenStack that includes carrier grade features that are defined later in this chapter.



0 Kudos
About the Author


Read for dates
HPE Webinars - 2019
Find out about this year's live broadcasts and on-demand webinars.
Read more
Read for dates
HPE at 2019 Technology Events
Learn about the technology events where Hewlett Packard Enterprise will have a presence in 2019.
Read more
View all