Managed Kubernetes

The future-oriented container orchestration

Containers in IT - the best-known representatives are the Kubernetes containers of the Google open source platform. They offer a solution to a problem that is all too common in IT operations: running software reliably and in a standardised way - regardless of the deployment target. Before containers, virtual machines (VMs) were the primary method used to run many isolated applications on a single server. VMs require a hypervisor host that abstracts resources.

Containers, however, do not need their own operating system. The container engine provides access to the kernel of the host operating system (host OS). Independent microservices are made available in the containers. Each individual service can be set up, implemented, scaled and renewed separately. Containers are therefore urgently needed for digital transformation and are faster, more portable and easier to update than VMs.

Advantages of using containers?

• Speed: 

  Unlike virtual machines, which usually take several minutes to start up, containers do not need to wait for the operating system to start and are ready within seconds.

• DevOps: 

  The speed, narrow footprint and resource efficiency make containers ideal for use in automated development environments and enable optimization of the CI/CD pipeline.

• Availability

  Containers package only the app and its dependencies. This enables easy operation on a wide variety of platforms.

• Scalability:  Containers are usually small as they require no separate operating system, such as virtual machines, and thus scale arbitrarily.
  
• Consistency:  Because containers hold all dependencies and configurations internally, they ensure that developers can work in a consistent environment, regardless of where the containers are deployed.
  
  

Google's open source Kubernetes offering

K8s explained simply - Kubernetes infrastructure

Kubernetes (abbreviated k8s) is an orchestration software. It allows software developers and IT administrators to organize, manage, monitor and automate the deployment, scaling, operation and administration of container environments. This management environment coordinates the computer, network and storage infrastructure. All of this is done on behalf of user workloads.

K8s is also a portable, extensible open-source platform and has a large, rapidly growing ecosystem. Kubernetes services, support and tools are widely available. Various IT infrastructures can be used, e.g. local hosters, AWS or Azure. In 2014, Google made the Kubernetes project freely available to everyone after the tech giant gained over a decade of experience running workloads in large-scale production. So container cluster management becomes a necessity especially for more complex projects with a large number of containers distributed across multiple servers or virtual machines.

Insights for Google Kubernetes Architecture

Pods: Pods are the smallest unit that can be deployed with Kubernetes. Each Pod encapsulates one or more containers, with one pod usually housing one container running one application or microservice. If pods contain multiple containers, however, it is often because they need to access the same resources.  These containers the share the host, storage, and network within the pod.

Node and Cluster: Pods run on nodes, i.e. virtual or physical machines, which in turn are grouped into clusters. Each node has the services needed to run pods. A cluster has control units called Kubernetes Masters, which are responsible for managing the nodes.

Project Borg and the Kubernetes story:  Containerized workloads have been around for quite some time. For over a decade, Google has been using them in their internal production. Whether it's service jobs like web front-ends, infrastructure systems like Bigtable, or batch frameworks like MapReduce - Google mostly uses container technology. Whether it's service jobs like web front-ends, infrastructure systems like Bigtable, or batch frameworks like MapReduce - Google mostly uses container technology. Today, we know that Kubernetes is directly based on Borg, and many of the Google developers working on Kubernetes were previously involved in Borg. Four central Kubernetes functions have emerged from the project: Pods, Services, Labels, and IP-per-Pods. By releasing the functionalities of the cluster manager, not only open-source projects within Kubernetes were created, but also publicly accessible hosted services via the Google Container Engine. Google is returning Kubernetes to the community "CNCF" and at the same time, wants to advance container technology through swarm intelligence.