We all know the hassle of working with more than one operating system and having the inevitable problem… yes you called it, failure of integration! There is possibly nothing more frustrating than this and in light of the fourth industrial revolution, many companies are expanding their IT infrastructure to support the substantial number of users hitting their servers. Although, just how much infrastructure is a company supposed to add when they can’t predict their scalability and how costly could this get? This is where your lifesaver and new best friend comes in – the innovation that is Docker!
What is Docker?
We know you’re dying to know, well Docker is a software development platform tool that makes it easy for us to develop and deploy applications inside neatly packaged virtual containerized compartments called containers. Docker is revolutionary in that it allows applications to run on any operating system (OS) environment, regardless of whether it’s Microsoft, IOS, MySQl, WordPress etc. These containers are portable and can simulate multiple instances of an OS/application on the SAME host machine or server. Containers can be deployed to just about any machine without any compatibility issues, meaning your software stays system agnostic, making software easier to use, less work to develop and ultimately way easier to maintain and deploy. It’s a game changer! Not only does Docker make it more efficient than the schlep of adding more infrastructure, since computing resources are taken advantage of, Docker is also horizontally scalable which allows for nodes/servers to be easily added, removed, stopped and started without affecting each other or the host machine, making it cloud-friendly.
How it works
Okay let’s get into the nitty-gritty! The Docker engine is installed on a host machine, this consists of a Docker Client, Rest API (Rest meaning Representational State Transfer – this is a set of rules that developers follow when they create their Application Programming Interface (API) as well as the Docker server (Docker Daemon). The Docker client runs commands which are translated by the Rest API, the commands are sent to the Docker Daemon which then checks the requests and interacts with the OS to create and manage the containers we spoke about earlier. Docker can be run using a command line tool and the Docker Daemon is run on the host machine.
The user then builds Docker images, this is a template consisting of instructions that create Docker containers, these images are stored in a Docker registry making the images accessible to other users. One of the best and key strengths of Docker is its ability to run numerous containers on the very same infrastructure and share the OS between each container without the containers knowing about each other, which is awesome.
Now that all the in-depth nitty-gritty talk is over, let’s talk about best practice.
When using Docker and containerization there are a few best practices to keep in mind for smooth sailing:
Firstly, the user must ensure that only one application is run in a container as a container should have the same life cycle as the applications that it hosts, in other words – when the application terminates, the container should terminate too. If a container consists of more than one application, Docker will not be able to tell if one of the applications is unresponsive making the container unhealthy, preventing the container from restarting automatically.
Secondly, the Docker build cache should be optimized. Basically, images in Docker are built layer by layer and because of this Docker saves the day again by trying to reuse layers from previous builds to minimize costly steps. It’s important that the Docker build cache is compiled in such a way that build steps that change often are at the bottom of the cache. This really helps as Docker is then able to use its build cache. Keep in mind it only does so if previous build steps in the cache have been used in the image.
Lastly, it’s important that users carefully consider whether to use private or public base images. Using publically provided base images can make the building process a lot quicker which is great at first, however if many changes need to be made to the base image – it’s recommended the user opt to build their own base image, which is better in the long run.
Following these best practices will allow Docker to be used to its full potential. Docker is here to save you time and money, it is efficient, smart and once you start using it, you’ll wonder why you took so long to invest in it.
Are you ready to take this revolutionary step? We’re ready to help you every step of the way. Contact us on web@piidigital.co.za today!
