Run-your-own or managed Kubernetes? Why Microsoft AKS may be the right orchestration platform for you

Container technology is triggering significant operational change in the cloud-computing era. It improves agility, streamlines development and deployment operations, increases scalability, and optimizes resources, but choosing the right container orchestration layer for applications can be a challenge. Should you manage it yourself or rely on a PaaS solution with a cloud-service provider? This question and more need to be addressed as part of your cloud strategy.

The Kubernetes open-source container- and process-orchestration system can deploy applications that are running either in containers or as processes. It is the standard container orchestration tool, and it is available for enterprises to manage on their own or is offered by cloud service providers like Microsoft as Kubernetes-as-a-Service.

Run-your-own-Kubernetes versus managed Kubernetes

Although many organizations decide to run the orchestration on their own, this means manually handling processes like lifecycle management, scaling, and scheduling, which requires a lot of time and investment.

Instead, leveraging Kubernetes services like Microsoft Azure Kubernetes Service (AKS) means these processes can be handled automatically. In fact, except when an organization requires a high degree of customization because of versioning at the master-node level, managed container orchestration services have significant advantages over running your own Kubernetes, including speed, ease of operations, and cost.

Managed Kubernetes with Microsoft AKS

Let’s explore some of the other benefits of AKS through the lens of common AKS deployment scenarios.

·       Easy migration of existing applications to containers: If an organization already has on-premise Kubernetes, it’s possible with AKS to leverage the existing CI/CD pipeline. It’s simple to migrate existing applications to a container and run within Microsoft AKS, with the environment making it easy to “lift and shift” existing applications rather than modifying code or re-factoring. Once you’ve migrated, AKS also allows you to control access via integration with Azure Active Directory and the SLA-backed Azure Service.

·       Simplified deployment and management of microservices-based applications: AKS simplifies the deployment and management of microservices-based architectures, which can be complex given the testing, debugging, and team collaboration that’s required. AKS streamlines horizontal scaling, self-healing, load balancing, and secret management. Also, for non-production-level AKS clustering, Azure Dev Spaces iteratively develops, tests, and debugs microservices so you don’t have to. Additionally, because you can leverage existing Kubernetes tools, you don’t need to reinvent the wheel when you decide to host your services in Microsoft Azure.

·       Rapid scaling: With AKS, you’re able to use virtual nodes to provision pods inside Azure Container Instance (ACI) in a matter of seconds, which means that AKS will run with just enough capacity for your average workload. As you run out of capacity in your AKS cluster, you can scale out additional pods in ACI without any additional servers to manage.

·       On-demand IoT device deployment and management: IoT scenarios can potentially involve hundreds to thousands of IoT devices. AKS provides IoT compute resources on demand, whether they’re running in the cloud or on-premises.

·       Machine-learning model training with AKS: Training of machine-learning models using large datasets is a complex and resource-intensive task. AKS supports common tools such as TensorFlow and Kubeflow to simplify the training of machine-learning models. This allows you to bring in your ML models in AKS clusters backed by GPU-enabled VMs.

·       Deploy applications quickly and at scale: By using Azure DevSecOps practices, you are able to deploy applications quickly at scale with enhanced security, developing and delivering code more rapidly using continuous integration and delivery. Policy enforcement within AKS helps you balance speed with security and ensures deep traceability.

The benefits of choosing a PaaS solution like Microsoft AKS to manage your Kubernetes versus running it yourself are significant. It helps you prepare for growth and business evolution and allows you and your teams to focus on containers and code without worrying about managing your own platform. If you’re interested in learning more about Microsoft AKS and how our team can leverage our experience to help you take advantage of its benefits, please contact me.

Jigar Pathak

Expert in Microsoft Azure, Public Cloud

Helping customers to move their business application to Cloud with consulting and training on cloud technologies.