The Linux Foundation Open Platform for Network Functions Virtualization (OPNFV) collaboration project is only a couple of months old -- it was launched at the end of September 2014 -- but there has already been some significant efforts put behind it by founding network equipment provider members. Ericsson, for instance, introduced at the end of October 2014 a "certification program" for OPNFV -- a sign that the vendor hopes the standards that OPNFV aims to develop will have some staying power.
OPNFV will be a useful addition to the efforts to get NFV into operators, insofar as it should help to reduce the fear that some of the benefits of virtualized network functions (VNFs) may be lost if proprietary technology stacks persist, or migration from physical to virtual environments proves too difficult. Compliance with open standards can be useful in such cases -- not just to vendors without comprehensive proprietary technology stacks, but to operators as well.
However, OPNFV is limiting its initial work to management of virtualized infrastructure and, in particular, to the definition of the interfaces between the virtual infrastructure manager (VIM) and the virtualized infrastructure itself, and between that infrastructure and the VNFs running on top of it. ETSI's NFV interface specifications are generally considered somewhat weak.
What OPNFV is not looking at, at least in the immediate future, is what happens higher up in the functional block diagram of NFV -- and there is just as much confusion over interfaces between NFV orchestrators and VNF managers, operations/business support systems (OSS/BSS) and models of services, VNFs and infrastructures. One issue is that different vendors with different portfolios of infrastructure, virtualization technology, management systems and high-level OSS/BSS advocate different ways of pulling things together -- orchestration -- depending on what they can provide and what their operator customers are asking from them.
Many large network equipment providers and the biggest OSS/BSS vendors are advocating a layered approach to orchestration, with an end-to-end service orchestration layer and domain-specific resource orchestration systems underneath. Given the likely persistence of hybrid physical/virtual environments in most operators' networks, this seems sensible, but it makes interfaces between orchestration and management systems even more important. It would be useful for OPNFV to consider increasing its scope to address this in the future.
The Heavy Reading Service Provider IT Insider, "Orchestration for NFV: The Big Players' View," examines the new vocabulary of orchestration in the context of NFV, and the battle for mind-share among industry bodies and the largest players in the network equipment provider, OSS and IT space. It considers how services built of VNFs are managed through their lifecycle, and the ways that the biggest vendors are responding to the challenge of delivering NFV management and orchestration (MANO) platforms to support their customers. Finally, it profiles 11 established vendors, describing their NFV orchestration strategies and relevant products, solutions and services.
— Danny Dicks, Contributing Analyst, Heavy Reading Service Provider IT Insider
Orchestration for NFV: The Big Players’ View, a 22-page report in PDF format, is available as part of an annual subscription (six issues) to Heavy Reading Service Provider IT Insider, priced at $1,499. Individual reports are available for $595. To subscribe, please visit: www.heavyreading.com/servsoftware.