The rise in network functions virtualization (NFV) is teaching IT management teams that there can be too much of a good thing. Every day IT organizations must combat virtual server sprawl in order to improve IT infrastructure that is shared by applications running on top virtual machines hosted on the same physical machine.
NFV popularity is making virtual server sprawl an issue that is likely to raise its ugly head on networks as well. Instead of having a limited number of physical networking appliances to manage, networking of the future will be defined by NFVs that shift networking functions into software that can run on industry standard x86 or other types of commodity processors.
Today most of those NFVs manifest themselves on a virtual machine. But it’s apparent that much of that NFV functionality is also going to appear in the form of containers such as Docker. In fact, network services will achieve a level of granularity that will enable dedicated application delivery controllers and even firewalls in the form of an NFV to be attached to specific application workloads.
As much potential as that future of networking might offer, however, managing granular sets of networking services will require a more sophisticated approach to network management. To achieve that goal, networking leaders have been driving investments in NFV management and orchestration (MANO).
At its core MANO consists of a NFV orchestrator responsible for on-boarding of new network services and virtual network functions (VNF); lifecycle management; global resource management; validation; and finally authorization of network functions virtualization infrastructure (NFVI) resource requests. Two other key components are a VNF manager that oversees lifecycle management of VNF instances and a Virtualized Infrastructure Manager (VIM) that controls and manages compute, storage, and network resources.
Of course, preventing NFV sprawl today will be extremely helpful in the future. There are already hundreds of types of NFVs. As this new class of software gets more widely deployed in 2017, organizations will discover that all the benefits associated with DevOps now apply to NFVs. Every time a new application workload gets deployed, an IT organization is going to be required to dynamically provision all the associated networking infrastructure required to enable it.
Rather than waiting for open source MANO technologies and standards to mature, IT organizations should put a MANO framework in place today that is capable of spanning both traditional IT environments as well as operations support systems (OSS) typically associated with Internet of Things (IoT) applications.
Over time IT organizations will have the option of replacing that framework with open source components as they mature. But knitting those components together will inevitably result in organizations falling further behind. In fact, the issues that a turnkey MANO framework addresses in the context of NFV deployments have already become pressing. IT organizations are under more pressure than ever to be agile.
Clearly, the least agile element of IT today is the network. NFVs are a big improvement in terms of enabling network agility. But the issue IT leaders need to keep in mind is that without some framework to manage them, those NFVs could quite easily become the cure that winds up being much worse than the original disease.