Overview of Docker
Docker is a platform that makes it comparatively easy to build, deploy, and run applications by using containers. Containers allow a developer to bundle an application with all of its components, including libraries and other dependencies, and deploy it as a single package. The developer may rest assured that the program will run on any other Linux machine, regardless of any customized settings that the machine might have that vary from the machine used for writing and testing the code, thanks to the container.
Docker is said to be quite similar to a virtual machine in several ways. Unlike a virtual machine, however, Docker allows programs to use the same Linux kernel as the system they're running on and only demands that applications be shipped with items that aren't already running on the host computer. This results in a major performance boost while also reducing the application's scale.
Docker is both free and open source. This means that everyone can contribute to Docker and expand it to suit their own needs if they needn't currently available features.
For whom is Docker designed?
Docker is a platform that benefits both developers and system administrators, and it's a key component of several DevOps (developers + operations) toolchains. For developers, this means they will concentrate on writing code rather than thinking about the machine on which it will eventually run. They can also get a head start by incorporating one of the thousands of programs already built to run in a Docker container into their application. Docker provides flexibility to operations personnel and can reduce the number of systems required due to its small footprint and lower overhead.
Security and Docker
Docker protects applications that run in a shared environment, but containers are not substitutes for taking adequate security precautions.
Dan Walsh, a computer security expert best known for his work on SELinux, discusses the importance of ensuring Docker containers' security. He also goes through the security features that are currently available in Docker and how they function.
Containers should be understood.
Containers are thought to necessitate three different types of software:
The term "builder" refers to the technology used to construct a container.
A container's "engine" is the technology that allows it to run.
"Orchestration" is a technology that is used to handle a large number of containers.
Containers have the advantage of being able to die gracefully and regenerate on demand. Containers are cheap to start, and they're built to appear and disappear smoothly, whether a crash kills them or because they're no longer needed when server traffic is poor. Since containers are intended to be ephemeral and to spawn new instances when required, it's anticipated that monitoring and management will be automated rather than performed by a person in real-time.
Linux containers have ushered in a huge shift in high-availability computing. There are numerous tool sets available to assist you in running services (or even your entire operating system) in containers. According to the Open Container Initiative (OCI), an industry standards group aimed at encouraging creativity while preventing vendor lock-in, Docker is one of several options. You can choose from various container toolchains thanks to the OCI, including Docker, OKD, Podman, rkt, OpenShift, and others.
If you plan to run services in containers, you'll almost certainly need tools to host and manage them. Container orchestration is the term for this. Kubernetes is a container orchestration system that works with a number of container runtimes.
What makes Docker special?
Docker has a simple workflow for moving an app from a developer's laptop to a test environment and finally to production. When you look at a realistic example of packing an application into a Docker image, you'll get a better understanding of it.
Do you know it takes less than a second to start a Docker container?
It is extremely fast and can run on any host with a Linux kernel that is compatible with it. (It also works for Windows.)
For image storage, Docker uses a Copy-on-write union file system. When a container is modified, only the changes are written to the disc using the copy on the write model.
You'll have optimized shared storage layers for all your containers if you use Copy on Write.
Core Architecture of Docker
The Docker architecture and its related components will be discussed here. We'll also look at how each part of Docker interacts with one another.
Since its development, Docker's architecture has undergone a few changes. Docker was developed on top of LXC and now has undergone several major changes:
In 2014, Docker switched from LXC to libcontainer, resulting in runc, a command-line interface for spinning up containers that conform to all OCI requirements.
containerd – In 2016, Docker split its container management component from containerd
OCI stands for Open Container Initiative, and it is an open industry standard for container runtimes and requirements.
Docker had a monolithic architecture when it first debuted. It is now divided into the three components mentioned below.
Docker Engine is a software platform that allows you to create
(dockerd)
Docker-containerd is a docker-based container system (containerd)
Runc-docker (runc)
Docker and other major corporations have agreed to contribute to the creation of a shared container runtime and management layer. As a result, containerd and runc are now part of the Cloud Native Foundation, with contributions from around the board.
Let's take a closer look at each Docker component now.
Docker Engine
It is a software platform that allows you to create the docker daemon, an API GUI, and the Docker CLI that make up the Docker engine. Docker daemon (dockerd) is a system utility that runs continuously. It is in charge of creating Docker files.
Dockerd uses the docker-containerd APIs to handle images and run containers.
Docker-containerd (containerd) is a system daemon service that downloads Docker images and runs them as containers. It makes its API available so that the dockerd service can send commands to it.
Docker-runc
The container runtime, or runc, is in charge of building the namespaces and cgroups that a container needs. The container commands are then executed inside those namespaces. The OCI specification is followed in the implementation of the runc runtime.
Docker's main components have been seen. However, there are additional components required to build, ship, share, and run docker containers.
Let's have a glance at the most important Docker components in a Docker ecosystem.
Components for Docker
The four components that make up Docker are as follows:
- Docker Client
- Docker Images
- Docker Daemon (dockerd)
- Registries for Docker
- Containers in Docker
Docker is built on a client-server model. The Docker daemon (dockerd) or server is in charge of all container-related operations.
The Docker client sends commands to the daemon via the CLI or REST API. The Docker client can run on the same host as the daemon or any other host.
The docker daemon listens to the docker by default.
UNIX socket, abbreviated as a sock. If you need to access the Docker API from a remote location, you'll need to expose it over a host port. Running Docker as Jenkins agents is one such use case.
You can use docker if you want to run docker inside docker from the host machine's sock
Docker Images
Docker's fundamental building blocks are images. To run a Docker container, you'll need an image. The OS libraries, dependencies, and tools required to run an application are contained in images.
For container creation, images with application dependencies can be prebuilt. If you want to run an Nginx web server in an Ubuntu container, for example, you'll need to create a Docker image that includes the Nginx binary as well as all of the OS libraries needed to run Nginx.
Docker has a Dockerfile concept that is used to create the image. A Dockerfile is a sort of text file in which each line contains one command (instruction).
A layered structure is used to organize a docker image. Every command in a Dockerfile becomes a layer in an image. A container is the image's topmost writable layer.
Every image begins with a base image.
For instance, you must be able to use an Ubuntu base image to create a new image that includes the Nginx application. A parent image or an image created from a parent image can be used as a base image.
You might wonder where this parent image (base image) came from. To create the initial parent base image, there are docker utilities available. It essentially bakes the required OS libraries into a base image. You won't have to do this because all Linux distributions' official base images will be provided.
The running container uses the image's top layer, which is writable. The image's other layers are read-only.
Docker Registry
It is a Docker image repository. You can exchange photos using the Docker registry. It serves as the main storage location for Docker files.
A registry may be open to the public or kept private. Docker Hub is a hosted registry service provided by Docker Inc. It gives you the ability to upload and import photos from a single location.
Note: Unless you define a custom registry in Docker settings, docker looks for images from the public Docker hub by default when you install it.
All of your photos are accessible to other Docker hub users if your repository is public. In Docker Hub, you can also build a private registry.
Docker hub works similarly to git in that you can create images locally on your desktop, commit them, and then push them to the Docker hub.
Set up your own docker registries instead of using the public Docker hub when using docker in enterprise networks/projects. Container registry services are available from all cloud providers.
Docker Container
It is Docker's execution environment. Photos are used to make containers. It's an image's writable layer.
This is how a ubuntu-based image looks when you try to relate image layers and a container.
You can bundle your applications in a container, commit it, and use it to create more containers using the golden image.
Containers may be started, stopped, dedicated, and terminated in a variety of ways. Any modifications made to a container will be lost if it is terminated without being committed.
Containers should be considered immutable objects, and making changes to a running container is not recommended. Only make adjustments to a running container for checking.
To create a tiered application architecture, two or more containers can be connected. Container orchestration tools like Kubernetes, on the other hand, have made hosting highly scalable applications with docker much easier.
Why is Docker so common, and why are containers on the rise?
Docker is famous because of the software distribution and deployment options it provides. Containers solve a lot of common problems and inefficiencies.
The following are the six key reasons for Docker's popularity:
1. User-friendliness
Docker's success stems in part from its ease of use. Docker is simple to understand, thanks to the numerous tools available for learning how to build and manage containers. Since Docker is open-source, all you need to get started is a machine with Virtualbox, Docker for Mac/Windows, or an operating system that supports containers natively, such as Linux.
2. More Rapid Device Scaling
Containers allow a lot more work to be done with a lot less computing power. The only way to scale a website in the early days of the Internet was to purchase or lease more servers. The cost of fame was equal to the cost of scaling up. Popular websites were casualties of their success, incurring tens of thousands of dollars in new hardware costs. Containers allow data center operators to pack a lot of workloads into a small amount of hardware. Hardware that is shared results in lower costs. Operators have the option of keeping their earnings or passing the savings on to their customers.
3. Improved software delivery
Container-based software delivery can also be more effective. Containers may be moved around. Also, they are fully self-contained. Containers come with a separate disc volume. As the container is built and deployed to various environments, the volume travels with it. The container includes all program dependencies (libraries, runtimes, and so on). If a container works on your rig, it'll work in a Development, Staging, or Production setting as well. Containers can minimize the problems associated with configuration variation when deploying binaries or raw code.
4. Adaptability
Containerized systems have greater flexibility and resilience than non-containerized applications. Hundreds of thousands of containers are managed and monitored by container orchestrators.
Container orchestrators are extremely useful for handling massive deployments and complicated systems. Kubernetes, the most common container orchestrator at the moment, maybe the only thing more popular than Docker right now.
5. Networking that is described by software
Docker supports Software-defined networking. Operators may define isolated networks for containers using the Docker CLI and Engine without touching a single router. Developers and operators can create systems with complex network topologies by using configuration files to describe the networks. This is also a security advantage. Containers in an application may run in their own virtual network, with tightly managed ingress and egress paths.
6. Microservices architecture is becoming more common.
The proliferation of microservices has boosted docker's growth. Microservices are small functions that do one thing well. They're typically accessed via HTTP/HTTPS.
Usually, software systems begin as "monoliths," in which a single binary supports a wide range of system functions. Monoliths can become difficult to manage and deploy as they increase in size. Microservices decompose a device into smaller, self-contained functions that can be implemented independently. Containers make excellent microservice hosts. They're self-contained, simple to deploy, and efficient.
