What is a cloud-native network function (CNF)?

What's the difference between VNF and CNF?

CNFs aren't the only method of moving network services away from hardware appliances and into software. Virtual network functions (VNFs) are another method that can be used.

With a VNF, the same software that is used on a hardware-based network appliance is migrated to a virtual machine (VM). Thus, the only difference is that processing and port usage are handled within software, as opposed to hardware. This differs dramatically from CNFs that pick and choose the specific network services required and then run them within a containerized environment, such as Kubernetes, instead of a VM.

With CNFs, the benefit is that processing and memory allocation are only needed for specific services, and the services can be distributed across a network depending on where they are needed. Thus, efficiency, scalability and performance benefits can be gained with CNFs when compared with VNFs.

What are the benefits of cloud-native network functions?

The benefits of CNFs largely revolve around large organizations that span multiple geographic locations and require substantial network infrastructures. These types of businesses can make the most out of CNFs. That said, the benefits of CNFs include the following:

• use of commodity hardware -- no specialized appliance hardware is required;
• improved scalability and performance in large, distributed networks;
• microservices architecture that enables easier application programming interface, or API, integrations with other platforms for data collection and analysis;
• centralized management plane; and
• smaller data center footprint for improved energy consumption.

Cloud-native network function use cases

As stated previously, taking advantage of the primary benefits of CNFs are best when a network is large and dispersed across a large geographic region. Thus, public telecommunications carriers, internet service providers and cloud service providers are the first to use these types of distributed network functions. These businesses replace aging physical or virtualized network appliances and substitute them with containerized CNFs that require a fraction of the compute, memory and physical footprints.

As new cloud or telecom points of presence come online, these companies can deploy CNFs and other container services on much smaller platforms, enabling micro data centers to be deployed for edge computing purposes.

Enterprises that operate in hybrid and multi-cloud architectures are also looking to CNFs so they can easily deploy network services in public clouds that do not allow for physical appliances. In these cloud environments, enterprises can also avoid deploying multiple virtual server network appliances that can cost a lot of money over time. For these businesses, flexibility and cost savings are the primary drivers for adopting CNFs.