What is Docker and Understanding its Architecture

 Hello my dear readers , Greetings & welcome back!

Understanding Docker (Continued...)

Docker is an open source for building environment for applications to package with their relevant dependencies and to run independently regardless of the environment (Operating system).

It is written in GO programming language which uses several feature of linux kernal to use various features. Docker uses a technology called namespaces to provide the isolated workspace called the container.

Docker uses containerization technology to run application without OS installation in that container. It packages only applications and application dependencies in a container to run on top of any OS without pre-configuration of Hardware like CPU and Memory unlike in VMs we usually do.

What is a container?

A container is a runnable instance of an Application derived/deployed from an image stored in image repository. You can move, build or rebuild, modify or delete a container or application inside the container. 

Containers are actually light weight application instances which run on top of OS and actually share same OS for hardware utilization required by applications.

What is Container Host?

So, a container is a loosely coupled workload isolated from an other Application running on the same Operating system and that operating system whether installed in side a VM or on a physical computer system is known as container host. In the case of Docker, This is known as Docker Host as shown in the figure below 

What is Image repository?

An image repository is actually a place for storing different applications templates we can use to deploy in a container with deployment instructions. In the case of Docker, this repository is known as Docker Registry.

Mostly, an image is based on some other customizations which may require to run application on some specific environment requirements. For example, an image may be built on top of Ubuntu OS with Apache web server and an application running on top of that web server be installed with some dependencies which may require to be deployed along with application in a container.

These settings for applications to be deployed in a container are written in a declarative language to instruct Docker to deploy application. This declarative statements are normally written in a language which could be YAML or JSON.

Docker Architecture High-level 

Basically, it is divided into three parts

1. Client (Docker)
2. Docker-Host (dockerd / Docker Daemon)
3. Docker-Registry

1. Client (Docker)

when any user or administrator tries to connect to docker host it always uses docker client to communicate with container which is running in a docker host. In this way, a client process talks to docker host (dockerd) through which it communicates with container / application running inside the container. The client uses docker API to communicate with dockerd for container management. Example, docker run which communicates with docker host to initiate docker management environment.

2. What is Dockerd (Docker Host)?

The docker daemon (i.e. dockerd) recieves commands from client and runs or manages containers, images / image repositories, networks and volumes. A docker daemon can communicate with other daemons to manage services.

3. Docker Registry?

It stores docker images for applications to build / deploy in a container. Example is docker hub is a public docker image registry. Anyone can use this public repository and docker by default is going to look into this registry for images but you can have your private or own image registries as well.

When docker pull or docker run is being used then container images from repository are being utilized to be used as source of application delivery.

What is docker image?

Docker image is actually a read only template (of applications for Specific or generic OS based environments) you prepare to push in the repository for later use and deployment of applications in a container.

Usually, an image is based on another image, with some additional customization. For example, you may build an image which is based on the Ubuntu OS etc. To build your own image, you create a Dockerfile with a simple syntax for defining the steps needed to create the image and run it. 

Below is the complete high level picture for understanding docker containerization using basic docker architecture what explained above.

Stay tuned ...😉

Coming up next "Installing Docker"

Go back to the previous topic

 

About the author , Twitter, Linkedin

Contributors:

Shumail Ahmed (Colleague) 

Muhammad Usman Hashmi (Colleague)

What is Containerization and how is it different from virtualization - Part 2

  Continuing the same discussion of rebuilding a VM from the point it got stuck or destroyed. It always require sometimes to follow below steps

1. Re-install OS
2. Configure OS
3. Install Application
4. Configure its dependencies

which may take more time to restore the services. In this connection, we always focus on services high availability which involves more hardware resources dedicatedly available to provide service High availability as shown in below picture for an understanding

So, in-order to utilize hardware efficiently for a muti-instance of an application which is an important point to focus on, including manageability, lets discuss this topic with the concept of containerization.

Lets conclude the discussion which leading us to think about containers...

1. Easy Management of workload (VMs = OS + APP)

2. Efficient resource utilization (CPU, Memory, Storage, Network)

3. Improved service re-instantiation (No need to re-install OS if App corrupted)

4. Zero chances of Application configurational Issues in-case of App re-build and Config

above four points are not the only points, we are going focus on containerization. That's the starting point to discuss this technology in more details.

In a nutshell, if we compare virtualization and containerization then 

• In virtualization, we consolidate Hardware to efficiently utilize it by distributing Hardware resources amongst the consumers (e.g. VMs / workload)
• In containerization, we consolidate OS kernel (which runs on top of H/W) to efficiently distribute resources running under the same OS to the workload (e.g. Applications).

Below picture shows a high level difference in between both of the technologies

NOW, being on clear note, i am going to mention some product names to get this concept very clear for my valuable readers with examples in my blogs. Those products are VMware vSphere, Docker, Kubernetes (K8s), vCenter Server Appliance, Microsoft Windows and Linux.

Containerization 

In this concept of containerization, I am going to choose "DOCKER" as an example to explain containerization. Just like VMs which provide isolation to OS to run independently without any conflict with any-other OS running on the same Hardware, containers provide isolation for an Application to run independently without any conflict with any other application to run on the same OS and hardware. So, hyper-visors provide this level of isolation to VMs to run independently similarly Container-runtime (CRT) provide level of isolation to APPs to run independently. Though there are so many containerization software available out there in the market and "DOCKER" is one of them. If we define "Docker" then below is the definition 

"Docker is an open platform for developing, shipping, and running applications. Docker enables you to separate your applications from your infrastructure so you can deliver software quickly".

Understanding DOCKER (High-level Overview)

Stay Tuned ... 😊

Go back to Previous Post

 About the author , Twitter, Linkedin

What is Containerization and how is it different from virtualization - Part1

What is a Container?

Hi everyone, I have initiated an effort for my blog readers to understand containerization architecture. In this detailed series of posts, I'll try to focus on simplicity of the concept with clear picture of understanding. 

So, let's start this discussion with some common questions😊

What is Container or containerization?

The concept behind this fascinating buzz word is the same we usually interact within our daily routine life. So, lets discuss this concept first ... 

As with the meanings of the word "Container", it is obvious that there is "an environment which provide all necessary requirements for a software package to run without any conflict with other software using the same operating system resources".

So, we breakdown an application into smaller (Sub-set of) services which are known as Microservices. Now, what is microservices? Microservices is an architectural style that structures an application as a collection of services that are. Highly maintainable and testable. Loosely coupled. Independently deployable. Organized around business capabilities.

Need for Resource Isolation

If I explain containerization in simple words as i am trying to portray then below diagram is really helpful in understanding this concept. Lets talk about a simple computer system you operate in your daily routine life in the form of a Desktop / Laptop. This computer system has got two major things running inside as software

1. OS (Operating System)
   
2. Application

Operating system occupies all the hardware resources of that computer system (as you can see in above picture) through a set of software known as drivers and on top of Operating System (OS) there runs an application system for a particular purpose or activity.

For example, Microsoft word is an application we can use for documentation, Excel we usually use for Accountancy or calculous. But hey wait, Tell me one thing, Can i install Microsoft Office 2020/2019 on the same computer system where Microsoft office 2013 already installed (in parallel) ? and the answer would be a big "NO" because both of the applications though from the same vendor but require separate physical environment as provided by single computer system.

Now, question is who is providing a separate environment for this application? and the answer is Computer system (e.g. Laptop or Desktop) and resources are provided to application from OS installed or running in that computer system.

Evolution of the Virtualization

But in most of the individual computer systems more than 60% of the hardware resources (like CPU, Memory, Storage, Network bandwidth etc) are not utilized by the Application and left untouched throughout the lifecycle of hardware which realized (around 20 years back). And this was the time when Server Consolidation came into picture and many organization started thinking about this gray area to be addressed 

In this ERA of Server consolidation, one physical computer system can entertain multiple operating systems running on top of it with hardware resource abstraction in an isolated fashion. So each of the operating system has got an isolated environment run by a system to produce hardware abstraction layer for all of the running OSs which is called Hyper-visor.

These hyper-visors provided a layer of isolation for every OS to utilize hardware resources as per allocation model and this containerized approach for an Operating system is known as Virtual Machine.

What is Hyper-visor?

Is a piece of software just like operating system who occupies all of the physical hardware and produces hardware abstraction (i.e. CPU, Memory, Storage, Network) to the virtual machine as VM resources. Without going into details that how many types of hyper-visors are just for your information at a high level overview, there are generally two types of hyper-visors

1. Bare Metal Hyper-visors (Type-1) We are talking about this 😊

2. Hosted Hyper-visors (Type-2)

What is Virtual Machine?

A virtual machine is a software representation of a physical computer system. Made up of numerous files and folders managed by Hyper-visor. This idea of resource efficient utilization has made a remarkable change in Business and IT industry and a new dimension has just been opened up for IT technology.

The Turning Point

But the problem faced by IT industry at VM level when we need to maintain huge number of virtual machines with the same OS/APP posture. In this context, we need to focus on more robustness and efficacy for application lifecycle. Below points are so important to look ahead for containerization needs and requirements

1. If any application require high availability how much RPO (Recovery Point Objective) we can achieve / reduce and same for RTO (Recovery Time Objective)
2. For a VM to rebuild we need to 
1. Install OS 
2. configure OS as per App Requirement
3. Install App
4. Configure App as per Business needs

Above points needed a consideration for reasons to reduce man effort and increase Application reliability. 

Click Here for next part

 About the author , Twitter, Linkedin

Troubleshooting Minikube Installation in Microsoft Windows Environment

Troubleshooting Minikube first Installation in MS Windows Environment

This Document is created to help you out in troubleshooting common hurdles while installing minikube in your hosted virtual environment. So, during installation of minikube if you face any problem which is not letting you deploy your Kubernetes architecture in the form of minikube then cross check below steps.

• When you download minikube, add minikube.exe to c:\Minikube or any location you can mention in the system variable later on
• Configure system variables with the name of Minikube presenting its path with name you call in CLI including Kubectl tool location as well as shown in the fig below

1)     

Above Picture show paths I have configured in my environment for accessing minikube / K8s running in a VM.

Note: This may be required when you are using Kubectl tool from your laptop or desktop to access minikube setup from within vmware workstation.

Now, verify that hyper v is running inside guest OS (i.e. Windows 8 VM)

Alternatively, you can also run hyperv from powershell using this command below. Make sure you’re logged in as an administrator:

and use below powershell command to enable Hyper-V

Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V -All

You can read more about enabling hyperv here

Second Scenario or Situation

If during deployment of minikube, VM doesn’t start or raises any error for incomplete deployment.

• Go to virtual switch manager

• Create a New Virtual Network switch (Internal)
• Call the switch "Minikube"

• Go to Control Panel\Network and Internet\Network and Sharing Center

• Go to your vEthernet status (will be on the right side underneath “Access type:” Internet, in the “Connections:” list

• Go into properties, and then the sharing tab. Check that “Allow other network users to connect through the conputer’s Internet connection” and select “vEthernet (Minikube)” from the list.

After configuring above settings, open up elevated command prompt and the punch in below command

Run:

.\minikube.exe start --kubernetes-version="v1.8.0" --vm-driver="hyperv" --memory=1024 --hyperv-virtual-switch="Minikube" --v=7 --alsologtostderr

And check once again :) 

For detailed Installation Guide you can visit my YouTube Video

 About the author , Twitter, Linkedin

Minikube Installation in Microsoft Windows Environment

Installing Minikube in Microsoft Windows Environment

Minikube Installation is quit simple and there are numerous resources available to have better understanding about its initial setup.

I have already, prepared a detailed video for this installation on MS Windows 8 running as VM in VMware workstation 14 Pro. 

Many, people may have different requirement as per their requirements. Keeping this point in mind, you need to visit URL to access minikube User page

OR

You can download Minikube for Windows directly  

Once, downloaded then you need to execute this Installer to Install Kubernetes in your laptop / desktop / VM.

for more information you can access minikube Handbook . This link contains sufficient resources to have a better hands-on.

For any further assistance, leave a comment and I'll respond you on that :)

 About the author , Twitter, Linkedin