Alejandro García
Product Tech Lead en Whitestack
Increasing efficiency, reducing costs, and improving delivery speed are essential to stay competitive. As infrastructure and network functions become more complex, traditional maintenance windows are becoming less and less viable. The answer lies in DevOps methodologies, where automation and continuous integration are fundamental.
Kubernetes and CNFs
Kubernetes has become the de facto standard for container orchestration. Its features allow the coordination of multiple containers across a distributed infrastructure through a robust and scalable platform.
Some of the operational challenges that Kubernetes helps solve are the following:
Automatic scalability: Kubernetes’ architecture supports running workloads across multiple nodes. This enables faster horizontal scaling of applications by automatically managing the distribution of containers.
Application fault tolerance: Kubernetes proactively monitors workloads and detects when failures occur. If this happens, Kubernetes redistributes containers to other nodes, helping mitigate the effects of these failures.
Zero-downtime updates: The deployment model in Kubernetes natively supports the gradual implementation of changes. This allows new application versions to be rolled out without downtime, automatically coordinating traffic management between new and old containers.
The great success of Kubernetes has led to network functions being deployed using containers. This new generation of network functions, known as CNFs (Cloud-Native Network Functions), represents a completely cloud-native paradigm for deployment, management, and operations.
Conoce la solución ideal para tu empresa
The Operator Framework
The Operator Framework elevates the automation provided by Kubernetes to a new level. Operators are Kubernetes extensions that automate specific application management and configuration tasks. This includes both Day 1 activities (application installation and configuration) and Day 2 activities (re-configuration, backups, scaling, management, and more). The ultimate goal of an Operator is to encapsulate operational knowledge within the Kubernetes cluster itself.
Kubernetes Operators work through a declarative approach, where the desired state of a system is defined rather than step-by-step instructions. The Operator continuously compares these desired state configurations with the actual state to look for discrepancies. When detected, it triggers a reconciliation process, taking the necessary actions to align reality with what the user has specified. Since all of this is automated, this approach ensures that the system always moves toward the defined state, even in the face of unexpected failures.
All components of the Operator Framework are open source. This has encouraged the development and release of Kubernetes operators for various tasks and applications, such as databases, monitoring tools, and even utilities for physical infrastructure configuration.
GitOps and automation
GitOps is a methodology that uses a Git repository as the single source of truth for the desired state of infrastructure and applications. By combining GitOps with Kubernetes and Operators, a fully automated workflow is achieved, from code release to deployment and management in production.
The complete workflow is as follows:
A user makes a configuration change in a Git repository. This change can involve a version update, new network configuration, an increase in resources or application replicas, or anything else that can be represented as a configuration file.
When the change is approved and merged, a Kubernetes Operator (which is constantly monitoring the repository’s state) detects it, downloads it, and analyzes it.
The Operator identifies what changes the user made and applies them to the Kubernetes cluster, following its deployment model based on gradual updates.
Automating the lifecycle of CNFs through GitOps brings numerous benefits:
Increased operational efficiency: Automation reduces the need for manual intervention, freeing resources and minimizing human errors.
Reduced downtime: Updates and maintenance of CNFs can be performed continuously, without significant interruptions.
Greater agility and flexibility: Changes can be deployed quickly, allowing a more agile response to business needs.
Add-ons to GitOps on Kubernetes
While GitOps and Kubernetes provide a robust framework for automating and managing the lifecycle of network functions, there are other DevOps practices and methodologies that can be employed complementarily. These alternatives may be better suited depending on the specific requirements of each environment and organization.
Infrastructure as Code (IaC): Many network functions are still deployed as virtual machines on cloud platforms (such as OpenStack). The IaC approach allows defining and provisioning the virtual infrastructure of those network functions using declarative code. This facilitates the consistent and repeatable creation, modification, and destruction of environments. Additionally, it enables all the advantages of using code, such as versioning.
Configuration Automation: Configuration automation platforms allow centralized and remote management and configuration of equipment (servers, network devices, virtual machines, and applications). Configuration tasks are defined programmatically, enabling standardization and automation even across multiple servers.
Continuous Integration and Continuous Delivery (CI/CD) Pipelines: These pipelines automate the building, testing, and deployment of applications, aiming to enable a continuous and reliable workflow. Platforms implementing these pipelines typically integrate with Infrastructure as Code and configuration automation tools, providing a single management point for all these activities.
Automation and DevOps methodologies are shaping the future of operations, both for large-scale applications in general and for network functions in particular. By adopting the GitOps philosophy—the current pinnacle of the DevOps movement—organizations can transform their CNF management into a model with greater efficiency, agility, and operational reliability. The result is more competitive and resilient networks, capable of adapting to today’s demands with unprecedented speed.
Acerca de nosotros
Whitestack es una empresa líder en el despliegue productivo de soluciones basadas en tecnologías y código abierto, con un fuerte foco en la industria de telecomunicaciones.
¡Contáctanos para conocer más!
