The microservice architecture is another approach to developing a distributed application. This approach is suitable for the agility, scale, and reliability requirements of modern cloud applications. As we know, a microservice application is decomposed into separate components independently to work together to achieve the whole system.
In microservice architecture, you can scale out independently on the specific functionality rather than unnecessarily scaling out other areas of the application. So, you can scale the resources, such as processing power or network bandwidth, for a specific microservice. But what about sharing infrastructures with another microservice? In this chapter, we will discuss this challenge. Containerization comes into the picture to solve the problem of sharing infrastructure between microservices and allowing microservices to be more autonomous.
Containerization allows you to run your microservices in a completely isolated environment. So, according to the containerization approach, a container is an artifact in which a microservice and its versioned set of dependencies, plus its environment configuration, are abstracted as deployment manifest files. The container contains all infrastructure-related dependencies, environment variables, and configurations. It is packaged together as a container image. And this image is tested as a unit and deployed to a host operating system.
Containerization is nothing but it is a different approach to development and deployment of the microservices. The container is an isolated and runnable instance of an image. That image contains everything required to run an application. In the container, an application can run without using the resources of another container or host machine. You also have full control over a container to create, delete, move, start, and stop this container using a CLI, such as the Docker client. Containers can be connected to each other using a network. A container acts like a separate, independent, and isolated physical or virtual machine.
Although a container looks like a physical or virtual machine, the containers use the technology and concepts very differently from virtual machines. Although a container runs an operating system, it has a file system, and it can be accessed over a network, just like a virtual machine. Let's see the following diagram on virtual machine:
As you can see in the preceding diagram, Virtual Machines include the application, the required dependencies, and a full guest operating system. Hypervisor is a computer software that shares and manages hardware. Let's see the following diagram about Containers:
As you can see in the preceding diagram, Containers include the application and its required dependencies. Unlike virtual machines, Containers share the Operating System and underlying Infrastructure with other containers. These are running as an isolated process on the host operating system. Because Containers share resources, they require fewer resources than virtual machines.