As telcos continue to rollout 5G, they are adopting software engineering approaches, like DevOps and CI/CD to improve how they run the networks and innovate. Some are exploring the potential of specific practices, such as GitOps, to elevate their processes. Here we address what GitOps is, how it is different to DevOps, and why large telcos like Deutsche Telekom and Orange are interested.
Introduction
Telcos face a set of unique challenges in their virtualised network operations:
- Complex network configurations: Telcos have intricate network setups with various components and interconnections. These configurations must be efficiently and consistently managed and maintained, while system updates/changes are carried out.
- Scalability requirements: Variability in network traffic and customer demands needs to be accommodated. Infrastructure and applications must be scaled to meet these requirements – whilst also ensuring optimal performance and reliability.
- Rapid software deployment: Telcos need to deploy software updates, new features, and services swiftly to keep up with market demands and stay competitive. Traditional manual deployment processes are time-consuming and error-prone, hindering the ability to deliver new functionalities in a timely manner.
As telco networks move to the cloud and become more and more software-orientated, managing and maintaining those networks requires new processes and operating models derived from software engineering practices. DevOps is one such practice which merges development and operations activities to ensure a shared, collaborative approach to software development. GitOps is a newer concept which takes DevOps a step further by placing a version control and automated deployment tool – Git – at the centre of software development and network operations.
Why are we talking about GitOps?
In a recent STL research programme, several telcos mentioned GitOps as an effective approach that enabled them to meet the challenges above while delivering uninterrupted network services. This operational framework is in the spotlight for its ability to simplify software deployment, drive rapid innovation, and ensure the reliability and scalability of network infrastructure.
GitOps enables network operations teams to:
- track changes to network applications,
- collaborate effectively,
- automate network software deployments,
- maintain infrastructure consistency,
- achieve fast and more reliable releases.
This article explains the fundamental principles of GitOps and its practical implications for telcos.
What is GitOps?
GitOps is a method to manage, deploy and adjust the systems that help move data around in telco networks. It relies on Git, a version control systemthat acts like a detailed tracker for any changes made to these systems. It ensures that there’s always a clear record of system settings and the configurations that control how data flows through the network.
The ‘desired state’ of the network – what the configurations and settings should be – is recorded in Git. When engineers need to upgrade or alter the network, such as to boost its performance or roll out new customer services, they input these changes into Git. After that, an automated process steps in to align the actual network and the new instructions precisely to ensure the ‘desired state’ continues, reducing the chance of service interruption and eliminating the need for manual adjustments.
GitOps also introduces the concept of ‘immutable infrastructure’ to telecoms. This means that engineers don’t try to change the setup of network devices while they’re running – which would be like fixing an aeroplane mid-flight. Instead, they prepare a brand-new configuration which can then replace the old one in a seamless way.
For telco operations, where even a small amount of downtime can affect thousands of customers, GitOps offers a more secure and streamlined method for implementing network changes. It allows teams to collaborate on improvements, keep a record of who made each change, and quickly revert to a previous setup if something doesn’t work as expected. This leads to a more stable network service.
How is GitOps different to DevOps?
DevOps is a cultural philosophy that integrates software development (Dev) and IT operations (Ops) to shorten the development life cycle and provide continuous delivery. It encompasses a broad set of practices and tools to improve the software delivery process. DevOps integrates development, testing, and deployment into a more seamless workflow.
GitOps is a subset of DevOps practices that use Git as the main tool for managing all the stages of the DevOps pipeline (i.e. the development and deployment pipeline). GitOps uses Git for version control and pull requests for managing infrastructure and deployments, rather than a mix of different tools and processes.
Operational efficiency
GitOps enables consistent deployment of infrastructure and applications across different environments so telcos can configure systems accurately and reliably. It leverages automation to deliver efficiency benefits such as:
- faster provisioning,
- reduced manual errors,
- simplified rollback processes.
Enhanced collaboration and governance
GitOps promotes collaboration among network teams. The use of version control and pull requests in GitOps helps development and network operations teams to work together more effectively. They can jointly work on code changes, review each other’s work, and provide feedback. The GitOps operational framework fosters better communication and alignment between teams.
GitOps also plays a role in ensuring compliance and auditability. Telcos can enforce governance measures and maintain a clear record of changes and actions due to the centralised control and traceability provided by GitOps.
Continuous integration and deployment
The framework supports continuous Integration and deployment (CI/CD) practices that work like a smart conveyor belt for software changes. GitOps uses automation to apply rules and make changes swiftly when issues are detected. Automation in a CI/CD context has several advantages:
- reduced manual effort,
- fewer errors,
- faster and more reliably timed software releases.
The ability to roll out updates and new features quickly – without impacting the infrastructure desired state – improves time-to-market and customer satisfaction. GitOps enables telcos to do this with a lower human resource requirement.
Are telcos using GitOps?
Deutsche Telekom
In response to the complexity posed by 5G, where software takes centre stage in network operations, Deutsche Telekom has shifted from traditional hardware-based solutions to embrace software in containers managed by Kubernetes.
Given this change in approach, GitOps plays a pivotal role, allowing Deutsche Telekom to store both code and configuration information in Git repositories, ensuring a unified source of truth. GitOps simplifies software delivery, enhancing consistency and reliability.
Deutsche Telekom has partnered with Weaveworks to implement GitOps at scale. They’ve adopted Flux, a software agent that continuously monitors and aligns production applications with their Git versions.
By embracing GitOps and Kubernetes, Deutsche Telekom is not only meeting the challenges of 5G but also gaining better control, reducing costs, and accelerating deployments.
Orange
When Orange began its 5G rollout, they faced a complex task beyond just delivering lower latency. Rapid deployment and consistent infrastructure changes across the expansive 5G network required efficiency and standardisation across different regions.
In response, Orange adopted GitOps as its operational model. By using GitOps, Orange could effectively manage both infrastructure and application changes. As Git is a tool familiar to their developers, it allowed them to methodically track alterations and confirm that the actual deployments aligned with the intended configurations, ensuring a uniform approach across different regions.
To further improve its deployment strategy, Orange used Kubernetes to manage containerised applications. With Kubernetes, Orange could customize its 5G service deployments to the unique demands of each region. In essence, GitOps acted as a blueprint, setting the desired state of the software and infrastructure for 5G. Meanwhile, Kubernetes was used to actually deploy and manage this infrastructure or “blueprint”. Each deployment unit, or “container”, managed by Kubernetes can be tailored for different requirements which makes it particularly useful for customising 5G deployments.
What are the key takeaways for telcos?
It’s likely that traditional methods of network management will continue to be shelved for more dynamic and agile solutions. The power of GitOps’ code versioning, change tracking, and efficient deployment capabilities makes it a useful tool for simplifying complex deployments and infrastructure consistency.
Orange and Deutsche Telekom are good examples of how adopting software-driven approaches like GitOps and Kubernetes can facilitate the 5G rollout.
Going forward, telcos should be prepared for the following:
- Holistic integration: GitOps and Kubernetes work best when viewed as interconnected components of a unified system.
- Customisation as standard: Adapting deployments to regional demands is essential, emphasizing the versatility of Kubernetes.
- Continuous evolution: The shift towards these kinds of software methodologies will only gain momentum with the adoption of 5G and beyond.
- Collaboration: Partnerships, as exemplified by Deutsche Telekom and Weaveworks, allow the marriage of joint expertise for innovative solutions.
Download this article as a PDF
Do you want to know more about our research in this area?
Telco skills in demand: key trends in hiring patterns
Telcos require access to new skill sets to manage transformed networks and operations, as well as to offer and scale differentiated services to customers. This article identifies which skills are in-demand in the telecoms industry and leverages LinkedIn data to illustrate how hiring patterns are changing.
Navigating digital identity: understanding what it is and its importance for telcos
Digital identity solutions constitute a promising revenue stream for telcos, yet few have been able to fully capitalise on this opportunity. So why did the coordination between SK Telecom, KT, and LG Uplus prove so successful?
Telecom KPIs: What do they say about the industry
STL Partners investigates the strategic objectives and associated metrics telcos report on: Do telcos still reference the same performance metrics? Or are they reflecting industry change? How are telcos taking stakeholders with them on their change journeys?