Are containers the end of virtual machines?

For quite sometime now virtual machines (VMs) have gained a lot of traction. The major reason for this trend was IT industries were totally convinced about the fact that instead of having a huge room filled with servers, it is better to deploy all your workload on a single piece of hardware. There is no doubt that virtual machines have succeeded as they save a lot of cost and work pretty well making failovers easier.

In a similar sense when containers were introduced they received a lot attention and have recently gained even more popularity amongst IT organisations. Well, there are a set of considerable reasons for this buzz; they are highly scalable, easy to use, portable, have faster execution and are mainly cost effective. Containers also subside management headaches as they share a common operating system. With this kind of flexibility it is quite easier to fix bugs, place update patches and make other alterations. All-in-all containers are lightweight and more portable than virtual machines.

If all of this is true, are virtual machines going extinct?

Well, for this answer you will have to deep dive into the complexities of both worlds.

How Virtual Machines work?

A virtual machine is an individual operating system installed on your usual operating system. The entire implementation is done by software emulation and hardware virtualization. Usually multiple virtual machines are used on servers where the physical machine remains the same but each virtual environment runs a completely separate service.

Consider a Ubuntu server as a VM and use it to install all or any service you need. Now, if your deployment needs a set of software to handle web applications you provide all the necessary services to your application. Suddenly, there is a requirement for an additional service where your situation gets tighter, as all your resources are preoccupied. All you need to do is, install the new service on the guest virtual machine and you are all set to relax.

Advantages of using virtual machines

• Multiple OS environments can run simultaneously on the same physical machine
• Easy to maintain, highly available, convenient recovery, and application provisioning
• Virtual machines tend to be more secure than containers
• Operating system flexibility on VMs is better than that of containers

Disadvantages of using virtual machines

• Simultaneously running virtual machines may introduce an unstable performance, depending on the workload on the system by other running virtual machines
• Hardware accessibility becomes quite difficult when it comes to virtual machines
• Virtual machines are heavier in size taking up several gigabytes

How Containers work?

You can consider containers as lightweight, executable packages that provide everything an application needs to run and function as desired. A container usually sits on top of a physical server and its host OS allowing applications to run reliably in different environments by subtracting the operating system and physical infrastructure.

So where VMs depend totally on hardware we have a new popular kid in town that requires significantly lesser hardware and does the task with ease and efficiency. Suppose you want to deploy multiple web servers faster, containers make it easier. The reason for this is, as you are deploying single services the containers require lesser hardware compared to virtual machines. The benefit of using containers does not end here. Docker, a popular container solution, creates a cluster of docker engines in such a way that they are managed as a single virtual system. So if you’re looking at deploying apps with scale, and lesser failovers your first preference should be containers.

Advantages of using Containers

• You can any day add more computing workload on the same server as containers consume less resources
• Servers can load more containers than virtual machines as they are usually in megabytes
• Containers makes it easier to allocate resources to processes which helps running your applications in different environments
• Containers are cost effective solutions that help in decreasing both operating and development cost.
• Bug tracking and testing is easier in containers as there isn’t any difference in running your application locally, or on test servers, or in production
• Development, testing, and deployment time decreases with containers

Disadvantages of using Containers

• Since containers share the kernel and other components of host operating system it become more vulnerable and can impact security of other containers as well
• Lack of operating system flexibility. Everytime you want to run a container on a different operating system you need to start a new server.

Now coming to the original question.

Are containers worth it? Will they eliminate virtualization entirely?

Well, after reading this article you must have already guessed the clear winner considering the advantages over disadvantages of each platform.

So, in virtual machines the hardware is virtualized to run multiple operating system instances. If one needs a complete platform that can provide multiple services then, virtual machines is your answer as it is considered a matured and a secure technology. If you're looking at achieving high scalability, agility, speed, lightweight, and portability, all this comes under just one hood, containers. With this standardised unit of software, one can stay ahead of the competition.

If you still have concerns over security and how a vulnerable kernel can jeopardize the cluster than you need, DevSecOps is your knight in shining armor. The whole idea of DevSecOps is to bring operations and development together with security functions. In a nutshell, everyone involved in a software development life cycle is responsible for security.

Kubernetes Containerd 1.1 Integration is now generally available

Top 7 DevOps tools in 2018

What’s new in Docker Enterprise Edition 2.0?