The innovation trajectory of the telecom industry has long been seen as incremental rather than revolutionary.

From the earliest coast-to-coast calls in 1914, to the ensuing 60 years, little has changed from an infrastructure standpoint. The introduction of fabric technology picked up the pace, and more recently, Network Functions Virtualization (NFV), which aims to move network functions running on dedicated hardware to virtual machines (VMs).

Now, the rollout of 5G services presents operators with both challenges and opportunities for dramatic change.

New service offerings will potentially open up entirely new lines of business. Our challenge will be to meet this demand with the flexibility to rapidly roll out new services to outpace the competition. Having the right infrastructure will be a key enabler for telecommunications to move from a "pipe" to a "platform", helping to lay the groundwork to become a direct platform for innovative, data-driven services.

While virtualization is still an important part of the 5G rollout, cloud native and containers are revolutionary directions for this new challenge.

VM era

NFV was, and is, a revolutionary idea: Essentially, anything that can be done on purpose-built boxes can also be done by virtual machines on commodity data center suites—servers, storage, and networking gear.

Due to the rise of VMware and other hypervisors, potential benefits of NFV include reduced cost and power consumption, the ability to easily share resources and quickly scale services up and down, and hardware portability.

All of this helps telecom providers build new 5G networks. But purpose-built boxes still face challenges to do one thing efficiently on networks that are too large to run manually, from performance, migration and coexistence, management and automation, and of course security and resiliency.

Not all of these challenges are storage-related, and many are being addressed. Operators continue to move more 5G functions into virtualized network functions (VNFs), but the network rollout will take time and is still a work in progress.

As part of this transformation, they need high-performance, reliable infrastructure that is not purpose-built. But "commodity hardware" doesn't mean "everything is the same," it refers to hardware that supports open standards. In fact, the real differences between storage vendors can be critical to the success or failure of NFV projects. Telcos should seek storage solutions that meet their key requirements for reliability, performance, automation and efficiency.

The future of cloud native

While virtualized network functions (VNFs) are currently transforming the telecom industry, the industry has gone one step further with cloud-native network functions (CNFs).

CNF uses containers, which is why you sometimes see CNF referred to as containerized network functions. At Pure, we prefer the term "cloud-native network functions" because "cloud-native" isn't really about where applications are deployed; it's about how. Cloud-native applications can run on bare metal in the data center, public cloud, or anywhere else. Just as VNFs provide telecom operators with many benefits, CNFs combine all of these benefits and more.

CNF helps address some of the challenges that still exist when using virtualization and VNFs. CNFs provide:

• Autoscale: Containers can easily spawn more containers. Whenever a process needs a new container, you can get the new one and will give up when done. Kubernetes just handles it all through code, so no operator intervention is required. This is a huge victory for large-scale telecom networks.
• Support for DevOps: DevOps revolutionized programming, allowing continuous integration/continuous delivery (CI/CD). This means operators can deliver new features, products and product updates faster.
• Incredibly fault-tolerant and fast restart: CNF is based on microservices, which can greatly reduce the operational and security risks of large-scale failures. Containers can be restarted almost instantly, and upgrades can be performed without downtime, allowing for automatic and fast rollbacks when needed. If a container fails, the system automatically generates a new container.
• Monitoring and Reporting: Given their scope, telecom networks are highly dependent on good monitoring and reporting. Fortunately, there are many tools in the Kubernetes world that do this, starting with Kubernetes itself. Other tools include Prometheus for monitoring and Grafana for reporting and alerting.

The benefits of migrating to CNF are many, but while application developers love Kubernetes and containers, they can be a challenge for unaccustomed infrastructure teams.

A wide range of data management tasks need to be performed: capacity management, data backup, disaster recovery, security, data migration, etc. The old method will not work because such methods do not support containers (eg old backup tools cannot successfully rebuild the container environment). A container-native solution is needed to deliver these expected enterprise data services to container-based applications.

Are infrastructure teams ready for this shift?

For KPN, one of the largest telecommunications companies in the Netherlands and a customer of the Portworx division, the best platform for KPN is the best of both worlds: software that simplifies and optimizes processes, while leaving enough space for configuring and optimizing individual clusters to meet specific needs of customers.

This is an advantage when talking to decision makers in different industries. Whether serving logistics customers, healthcare or transportation, everyone thinks the same when it comes to the performance of information systems.

For the telecom industry, the road ahead leads to the cloud. CNF is the next step in upgrading operator networks, paving the way for operator innovation, which also makes customers' lives easier and businesses more resilient and successful.