Containers in Cloud Computing

Back to: Cloud Computing Tutorials for Beginners and Professionals

Containers in Cloud Computing

In this article, I will discuss Containers in Cloud Computing and how they are different from virtual machines. Please read our previous article discussinf Virtual Machines in detail. In our previous article, we discussed what are virtual machines and how they allow us to run multiple applications on a single physical server. However, with each virtual machine that we create, there is a lot of overhead, both in terms of cost and maintainability. Let’s understand this with an example.

Virtual Machines and Multiple Applications:

Imagine that we have three applications, namely App 1, App 2, and App 3, and we want to run them all on a single physical server, which we can call Server 1. However, we cannot run these applications directly on the server hardware. We need an operating system like Windows or Linux installed on the Physical Server. This operating system is called the Host Operating System, installed at the server level. To enable the running of these applications, we installed a software called Hypervisor on top of this host operating system. The Hypervisor is responsible for creating and managing virtual machines. So, to host the three applications, we create three virtual machines on top of the Hypervisor. We then host each application within each virtual machine, as illustrated in the image below.

As we discussed in our previous article, a virtual machine virtualises the hardware of the physical server. This means each virtual machine that is created, gets a slice of the server hardware. In our case we created 3 virtual machines, so each VM gets 33.33% of memory, disk space, processor and other server hardware resources.

Disadvantages of Virtual Machines in Cloud Computing

One benefit of virtual machines is that we can run multiple applications on a single physical server, effectively using the server’s resources, such as memory, disk space, and processor. However, it is necessary to create separate virtual machines for each application, which comes with a significant amount of overhead in terms of cost and maintainability. 

It is important to note that applications, i.e, App 1 to App 3 cannot directly run on a virtual machine. A virtual machine virtualises the server hardware. This means, in simple terms, a virtual machine is a slice of the underlying server hardware and an application cannot run directly on the hardware. So for a virtual machine to be able run a software application, it needs a full copy of its own operating system.

We have 3 virtual machines, which means we need to install 3 operating systems, one on each virtual machine. Usually, Windows or Linux. The operating system we install on a virtual machine is called the Guest Operating System. On top of this guest OS on the VM, we install our business application and any dependencies it might need. Installing Windows OS on one virtual machine and Linux OS on another is also possible. For a better understanding, please have a look at the following diagram.

Now, just imagine the, amount of server resources required to install and run 4 operating systems. So much of server resources like disk space, memory and processor time are wasted just to keep these VMs up and running. In addition to the wated server resources, each installation of the operating system also needs admin time. For example if there is an update or a security patch available for the OS, the admin needs to download and install it on all the VMs that has that operating system. In addition to this, we also need to pay for 3 operating system licenses.

Remember a virtual machine is like a computer with in a computer. It has it’s own OS. So for a virtual machine to be up and running it’s operating system must also be booted up. Booting up an entire operating system obviously consumes a lot of time. In our example, if we want App 1 to be up and running, VM 1 operating system must be booted up, in addition to booting up the host operating system i.e the operating system at the physical server level.

What are Containers in Cloud Computing?

A virtual machine virtualises the server hardware where as containers virtualises the operating system, i.e., it’s an abstraction at the operating system level. Multiple containers can run on the same machine and share the host operating system kernel. Unlike a VM, a container does not required it’s own operating system. 

So, Containers in Cloud Computing are a lightweight form of virtualization. They enable developers to package applications along with their dependencies into a single runnable unit, known as a container. This unit includes everything needed to run the application, such as the code, configuration files, libraries, and rruntime environment, ensuring that it can run consistently across different environments such as development, testing, and production. This technology is designed to solve the problem of “it works on my machine,” ensuring that software operates the same in different computing environments.

Understanding Containers By Comparing with Virtual Machines in Cloud Computing

Now, let us understand containers by comparing them with Virtual Machines. A virtual machine virtualizes the server hardware resources (memory, disk space, processor, and other server hardware resources), whereas a container virtualizes the operating system, i.e., it’s an abstraction layer at the operating system level.

It is possible to run multiple containers on the same machine, and they will share the host operating system kernel. Unlike a Virtual Machine, a container does not require its own operating system. As it does not require its own OS, it immediately saves disk space, RAM, and processor time. For a better understanding, please have a look at the following diagram.

The container packages the application code and its dependencies together. As you can see in the above image, if we have three applications to run on a single physical server, then we need to create three containers. Now, each application thinks it is running on a dedicated Operating System with dedicated server hardware, but all the container applications share the same host operating system and hardware.

In case of VMs, admin time is required to install security patches and keep up to date host operating system and the guest operating systems on all the VMs. No matter how many containers we have on our server, there is just one host operating system that demands admin time.

From the cost standpoint, in case of VMs, we have to pay for host and the guest operating systems of all the VMS, where as in case of containers, we have to pay just for the 1 host operating system.

A container is very small compared to a VM. Since a container does not have it’s own operating system to bootup, it can usually be brought online in a very less time compared to booting up and entire VM and it’s operating system.

Benefits of using Containers:

The following are some of the benefits of using containers:

• Isolation: Each container encapsulates an application and its dependencies, ensuring they run across different computing environments.
• Portability: Containers can run on any system that supports the container runtime and containers include everything needed to run an application which makes it easy to move applications between development, testing, and production environments.
• Scalability: Containers can be easily scaled up or down, making them ideal for microservices and modern cloud-native applications.
• Efficiency: They are lightweight and consume fewer resources compared to VMs, enabling higher efficiency and faster deployment.
• Reproducibility: Containers ensure that applications run the same way in any environment, reducing bugs related to environmental differences.

Use Cases of Containers:

Containers are particularly useful in the following scenarios:

Microservices Architecture: A company is transitioning from a monolithic application to a microservices architecture to improve scalability and maintainability. Containers allow each microservice to run in its own isolated environment with its dependencies, enabling independent development, deployment, and scaling of each service.

Continuous Integration and Continuous Deployment (CI/CD): A development team needs to automate the build, test, and deployment processes to release software faster and more reliably. Containers provide consistent and reproducible environments for CI/CD pipelines, ensuring that code runs the same way in development, testing, and production stages.

Hybrid and Multi-Cloud Deployments: A business wants to deploy its applications across multiple cloud providers and on-premises data centers for redundancy and flexibility. Containers offer portability and consistency, making it easy to deploy applications across different environments without compatibility issues.

Rapid Scaling: An e-commerce platform experiences fluctuating traffic, requiring the ability to scale applications up or down quickly. Containers can be easily scaled in response to demand, allowing the platform to handle traffic spikes efficiently and reduce resource usage during low traffic periods.

Environment Consistency: A development team faces issues where applications behave differently in development, testing, and production environments. Containers ensure environment consistency by packaging applications with all their dependencies, ensuring they run the same way across different stages of the development lifecycle.

Big Data and Machine Learning: Data scientists and engineers need consistent environments to run big data processing and machine learning workloads. Containers can package data processing frameworks and machine learning libraries, ensuring consistency and reproducibility across different environments and simplifying collaboration.

Container Orchestration

As the use of containers grows, especially in large-scale systems, managing them can become complex. Container orchestration involves managing the lifecycle of containers, especially in large, dynamic environments. Orchestration tools automate tasks such as deployment, scaling, networking, and load balancing of containerized applications. They are essential for managing complex applications that consist of multiple containers running across multiple hosts.

Popular Orchestration Tools:

• Kubernetes: The leading container orchestration platform, developed by Google, provides mechanisms for deploying, maintaining, and scaling applications. Kubernetes automates the distribution and scheduling of application containers across a cluster in a more efficient way.
• Docker Swarm: Docker’s native clustering and scheduling tool for Docker containers, which uses the standard Docker API and networking, making it easy to integrate into existing Docker environments. It allows Docker containers to be managed and scheduled on a cluster of nodes.
• Apache Mesos/Marathon: A high-availability cluster manager that provides efficient resource isolation and sharing across distributed applications or frameworks. Marathon is a Mesos framework for long-running applications.
• Amazon ECS: A highly scalable, high-performance container management service that supports Docker containers and allows you to run applications on a managed cluster of Amazon EC2 instances.

Differences Between Containers vs Virtual Machines (VMs):

Virtual Machines (VMs) and Containers are both technologies used to create Isolated Environments for running applications, but they operate in different ways and serve different purposes. Let us do a detailed comparison between Virtual Machines (VMs) and Containers:

Architecture:

• Virtual Machines: Virtual Machines virtualize the hardware, meaning each VM runs its own full-fledged OS. They are managed by a hypervisor like VMware ESXi, Microsoft Hyper-V, or KVM (for Linux), which sits between the hardware and virtual machines. VMs have significant resource overhead because each VM runs a separate OS.
• Containers: Containers use OS-level virtualization to create isolated environments that share the host OS kernel. Each container packages the application and its dependencies, but not the entire OS, making them lightweight and faster to start.

Performance:

• Virtual Machines: VMs tend to have slower performance due to the need to run multiple full operating systems, which consumes more CPU, memory, and storage. Boot times are longer compared to containers.
• Containers: Generally, offer higher performance due to the absence of a full OS and the efficiency of sharing the host OS kernel. They start up almost instantaneously and use fewer system resources than VMs.

Security:

• Virtual Machines: Offer a higher level of security due to complete isolation. Even if a VM’s OS is compromised, it does not affect other VMs on the same host.
• Containers: While containers offer some level of isolation, they are not as isolated as VMs since they share the host OS kernel. Vulnerabilities in the kernel can potentially be exploited to affect all containers on the same host.

Management and Maintenance:

• Virtual Machines: Management can be more cumbersome because each VM is a full OS, requiring traditional administration tasks like patching, security, and updates.
• Containers: Easier to manage, especially with orchestration tools like Kubernetes and Docker Swarm, which automate deployment, scaling, and management of containerized applications.

Use Cases:

• Virtual Machines: Suitable for running different operating systems on the same hardware, legacy applications, and scenarios requiring strong isolation.
• Containers: Ideal for microservices architecture, DevOps, CI/CD pipelines, application modernization, and scenarios requiring rapid scaling and portability.

In the next article, I am going to discuss the Advantages and Disadvantages of Cloud Computing. I try to explain Containers in Cloud Computing in this article, and I hope you enjoy this article.