Introduction to containers

Businesses are constantly looking for newer and quicker ways to deploy applications and keep businesses running. These requirements lead to software development and support teams to find solutions for saving money and time. Adopting an approach of using containers for rapid deployments leaves more time for software deployment and reduces time spent on creating and configuring environments.

Containers are specialized packages of software that bundle application software with associated configuration files, libraries, and dependencies that are required to enable an application to run. Containers also contain the kernel for an operating system (OS) and don’t require the whole OS in itself, which allows for larger-scale deployments than what VMs could achieve using the same resources. The reason for this is that the OS consumes resources, such as storage, memory, and processing power. In the previous chapter, you learned about VMs and how they work by logically creating a machine using resources; we say that VMs abstract the hardware layer. Containers, by association, abstract the OS layer. When deploying a VM, you will still need to specify the type of OS you require— this is typically Windows or Linux. The exact kernel version won’t be required to make it work as the kernel will be packaged in your container, and you just need to concern yourself with the application code, libraries, and dependencies to get your application to work. Using this approach allows for easy scaling, and only a single OS is consumed on the host of all the containers for the OS type, such as Windows or Linux. Being more lightweight, containers become easier to manage and much faster to deploy and scale. They introduce a standardized and repeatable mechanism for packaging, managing, and deploying applications.

Prior to containers, the challenge was writing applications for various systems with different versions of dependencies that the developers couldn’t control. This leads to several issues that make solutions very difficult to manage. With containers, you can standardize the deployment with dependencies and enable a consistent experience for all users.

Azure container instances allow you to run containers in Azure without the requirement to deploy your own container host or VM.

Top Tip

Windows container hosts can run both Windows and Linux containers; conversely, Linux can only run Linux containers to date. Running Linux containers on Windows will require Hyper-V services and using a system called LinuxKit (essentially a Linux VM). When deploying to ACI, however, it is important to select the correct OS for your requirement, as Linux containers run on the Linux OS and Windows containers run on the Windows OS.