Docker: a fundamental platform for DevOps
Docker Container, might be one of the most essential building blocks in the field of DevOps currently. It acts as a platform, which uses OS-level virtualization to deliver software packages, known as Containers. It is a client-server architecture. It is a tool, that has helped in reducing the gap between Development and Operations team of a project. This tool helps developers, in creating, deploying and running containerized applications on various platforms. In this article, we’re gonna explore about Docker and DockerHub, along with that, we’re also gonna explore its benefits and applications.
Difference between Docker Image and Docker Container
Before learning about the Docker platform itself, we need to learn about these two terminologies: Image and Containers. We are essentially are going to working with Images and Containers on Docker.
So, Image is basically a file, comprising of many layers, that execute the source-code of the application. It is built from, complete and executable version of an application. Applications are stored in the form of images.
On the other hand, when we want to create an instance of these images, to run them on Docker, they’re converted into containers. So, by using OS-level virtualization (running multiple instances on same hardware, isolated from each other), Docker can run multiple containerized applications on machine, isolated from each other.
Working of Docker Engine
Containers (which act as an instance to image of application) works in very similar way to any virtual machine. However, instead of creating a whole virtual guest operating system, Container uses same Linux kernel, on which they’re running on, for running the applications. This gives significant performance boost, compared to Virtual machines as one doesn’t need make adjustments to the machine, to run a specific application and containers themselves do not put any load on hypervisor, due to them being light weight. By using OS-level virtualization, Docker can run multiple, light-weight containerized applications on machine, isolated from each other.
For creating, deploying and running Docker containers, one only needs to install Docker daemon. This Docker daemon is a service that runs on the host operating system, which helps in creation, deployment and running of Docker container. It helps in creation and management of Docker objects like Container, Images, Networks and Volumes. We can interact with Docker daemon, using command line interface (CLI), while daemon exposes itself to the CLI using REST API. CLI, REST API and Docker daemon are collectively known as Docker Engine.
Docker Registry and Docker Hub
For publishing and storage of our Docker images, we have two ways:
Docker Registry is a highly scalable, server-side application, which helps in storage and distribution of Docker Images. It is used to store your Docker Images onto your local machine. It could be used when, you strictly want to control the storage of your Docker Images, or you want to integrate storage and distribution of your image, in an in-house environment.
On the other hand, you could use Docker Hub. It is a cloud based online platform, where developers can share, test, store and distribute container images. One can create their own open-source public repository on Docker Hub, for others to use. It works in similar way, in which GitHub does. We can also create a private repository on the platform, to store an organization’s work (which may not be open-source). Docker Hub is basically a cloud-based implementation of Docker Registry.
Elimination of “It works on my machine!” excuse
We all know, applications are generally platform dependent. Sometimes, it may happen that, Development team of a particular application, might be using a different set of dependencies on their systems to Operations team. There may be a situation, where the application may be working perfectly fine on the machines of Development team, but may not work at all on the machines of Operations team. Hence, Docker can be very useful in this aspect. When any Docker image is created, it contains set of instructions, to run that application as a Docker container, on any platform or machine, irrespective of any host operating system installed on the machine running the Docker engine. Hence, in this way we are able to reduce the gap between the Development and Operations team. Hence, this why Docker is considered an essential platform, for the DevOps field.
Future scope of Docker
Since its launch in 2013, Docker has already seen immense success in recent times, due to its many benefits like platform independence, enhanced performance due to containers etc. Docker Hub has also been popular, where big developers have started sharing images of their popular applications like WordPress and MySQL. It has formed a great community of developers, who try to create, test and publish Docker images together on Docker Hub. What began with 1146 lines of code has today turned into a billion-dollar product. Docker has grown to a stage where a majority of leading tech firms have been motivated into releasing additional support for deploying containers within their products. Examples include Amazon integrating Docker into the Elastic Beanstalk system, Google introducing Docker-enabled ‘managed virtual machines’, and announcements from IBM and Microsoft with regard to Kubernetes support for multi-container environments.
A new concept has emerged, in which developers are able to run multiple Docker containers on various nodes of a Kubernetes cluster. Due to ease in deployment on cloud resources, and immense rise expected in cloud industry, Docker has still a long future ahead of it. Docker has laid out its goals in developing core capabilities and cross-service management and messaging.
