Kubernetes
Kubernetes is an open-source platform (Google opened it in 2014) that is a container management tool. The platform serves as a tool for monitoring workloads and services associated with the containers' operation.
Platform settings facilitate declarative customization and process automation. Now the platform is the centre of an entire ecosystem with its tools set and services.
Kubernetes allows to deploy and run applications in a microservice architecture. This brings several benefits to development teams:
- Resource Consumption Control: Automatically restart applications that consume unnecessary resources.
- Equal load sharing according to local infrastructure
- Flexible resource system when updating a cluster, for example, is inviting a new host cases
- Autorequests for load rebalancing between application versions
- It is easy for an application to migrate from one host to another in a resource shortage or failure event.
- Easily rollback and launch canary releases aimed at testing synergies between new patches and old code.
The platform has a self-service system and provides the ability to create hardware abstractions for platform users. For development teams, there is also a quick request-response option in supplying resources terms for levelling loads.
Kubernetes is also considered a more optimized solution since the containers use consumes resources (CPU, memory) a magnitude order more efficiently relative to VMs or hypervisors.
How Kubernetes works
The kubernetes building block is a cluster that is created from a physical or virtual machines array. The machines, in this case, play the nodes' or connecting nodes' role - Control Plane Node. The node stores a containers group, which in turn contain applications. Control Plane in this scheme determines the containers parameters and new traffic routes when containers change location.
The container runtime (for example, Docker) launches containers and monitors their work when, using Kubernetes tools, a specialist deploys them to the cluster nodes.