Telco cloud: what is that?

Tytus Kurek

on 21 October 2020

Telco cloud or a network function virtualisation infrastructure (NFVI) is a cloud environment optimised for telco workloads. It is usually based on well-known technologies like OpenStack. Thus, in many ways, it resembles ordinary clouds. On the other hand, however, it differs from them. This is because telco workloads have very specific requirements. Those include performance acceleration, high level of security and orchestration capabilities. In order to better understand where those demands are coming from, let’s start with reviewing what kind of workloads are telcos running in the cloud.

Telco workloads in the cloud

Have you ever been wondering how the telecommunications infrastructure works? You probably have not, but do not worry, you are not the only one. All we usually care about today is a stable Internet connection. Understanding how does it work is of secondary importance. However, behind a network socket or your Wi-Fi router, there is a massive infrastructure which provides this connection. It consists of thousands of interconnected services, including firewalls, base transceiver stations (BTS) for providing mobile connection, voice and data aggregation systems, etc.

Historically, all of those services used to be implemented in hardware. Nowadays, however, service providers are moving to software-based network services. The migration is driven by economical benefits resulting from better utilisation of resources in cloud environments. As software-based network services are implemented on top of virtual machines (VMs) or containers, service providers can simply run them in a cloud, benefitting from lower operational costs and improved agility. Such a telco cloud, however, must meet certain criteria before network services can be deployed on top of it.

Telco cloud under the hood

In order to implement a telco cloud, service providers can use either proprietary or open source technologies. Over the past few years, it has been concluded that for the open source telco cloud implementation OpenStack will be used as the basis. What makes the telco cloud different from an ordinary OpenStack cloud, however, are very specific features required by telco workloads.

Performance

Among various metrics, performance results are what telcos care the most. This is because telco workloads are network-heavy. They have to process up to 100 Gb per second. Thus, it is important that telco workloads achieve comparable performance results regardless of whether they are implemented in hardware or in software. This is challenging, however, as VMs usually cause performance degradation. In order to solve this problem telco clouds implement a bunch of performance extensions, such as single-root input/output virtualisation (SR-IOV), data plane development kit (DPDK) or central processing unit (CPU) pinning. All of that allows software-based network services to achieve performance results comparable to those achieved by physical machines.

Security

Another important aspect is security. Telcos are known for being security-oriented. Thus, the telco cloud must provide a desired level of security too. Service providers usually achieve that by applying hardening on the operating system level. Hardening is a process of securing the system by reducing potential vulnerabilities to an absolute minimum. This is achieved by disabling unnecessary services, narrowing down permissions, closing open ports, etc. For obvious reasons, telco cloud is also deployed on-prem in most of the cases. The security team can later use standard technologies, such as packet inspection or data encryption to secure the telco cloud at each layer of the infrastructure stack.

Orchestration

Last but not least orchestration is what characterises the telco cloud as well. Although orchestration is a broader term in general, it is especially important in the case of telco workloads. This is because software-based network services are usually very complex. They consist of multiple interconnected components (network functions) which are often distributed across multiple substrates. Thus, having a tool which can arrange resources, deploy network services and maintain them post-deployment is important for service providers. Among various proprietary and open source solutions, an Open Source MANO (OSM) project has recently been getting momentum, enabling telcos with management and orchestration (MANO) capabilities.

Telco cloud on Ubuntu

Canonical is an established leader in the field of implementation cloud environments for telcos. Over the past few years, the company has successfully onboarded leading global and national tier-1 service providers like AT&T, BT or Bell on their open source NFVI platform based on Ubuntu Server, Charmed OpenStack and Charmed Ceph. With an increasing demand for cloud-native network services Canonical also stands by ready to offer Charmed Kubernetes as an extension of the underlying cloud platform. Finally, as workloads orchestration becomes the biggest challenge in the telco world nowadays, the company provides Charmed OSM to enable service providers with these capabilities.

To get in touch with Canonical with regards to solutions for telecommunications, click here.

To learn more, watch the webinar: “NFV, cloud-native networking and OSM: everything you need to know” or visit Canonical’s website.

kubernetes logo

What is Kubernetes?

Designed with economics in mind, Canonical’s solutions for telecommunications ensure ROI, providing first class quality at the same time.
Save costs by operating your infrastructure and applications the smart way, ensuring full automation from day 0 to day N.

Learn more about Ubuntu for telco ›

Newsletter signup

Select topics you’re
interested in

In submitting this form, I confirm that I have read and agree to Canonical’s Privacy Notice and Privacy Policy.

Related posts

CLI-only MAAS operation

MAAS provides a state-of-the-art User Interface (UI), which simplifies usage. But you may not know that MAAS also has a robust Command-line Interface (CLI),...

Exploring ROS 2 Kubernetes configurations

Kubernetes and robotics make a great match. However, as we have seen, robots running ROS 2 can be tricky to set up on Kubernetes. This blog series has...

Distribute ROS 2 across machines with MicroK8s

Introduction Our simple ROS 2 talker and listener setup runs well on a single Kubernetes node, now let’s distribute it out across multiple computers. This...