It’s a Long Way to the DevOps…

A Short Introduction

DevOps is a phrase of business software Development, used to signify a kind of agile relationship between development and IT operations.

The life cycle of traditional software development takes weeks to put work into production. DevOps seemed to improve this whole process by automating the infrastructure, the workflow and the performance of the Application.

The operations team (Ops) is responsible for diagnosing errors or problems caused by the changes. Therefore, the developers (Dev) must solve the problems and the circle starts again.

Packaging

We have a very big project with many applications written in AngularJs, JavaScript, C# and legacy code in SourceSafe. All this code is builded with a Team Foundation Server Build Definition and it´s output is the base of many installers that are created with a Jenkins´s Job using DotNetInstaller. Finally all these Setups are deployed in many servers by other Jenkins´s Job.

The DevOps manager is going by the idea of a predictable packaging process.

Why is this important? By implementing software in production, we take our system to a planned downtime. During this time, our customers can not access the application. Normally we do it during the night or during the weekend. If we do not package something necessary for deployment, it will fail and affect the production environment. The package allows us to reduce implementation errors and increase the speed of implementation.

• The standards allow us to support many solutions with the same knowledge.
• The standards provide a layer to improve the process and be creative without adding friction in delivery.
• The package of all the dependencies minimizes the changes of failures and allows us to have a similar production and development environment.

This package includes MSI setups, prerequisites, tools, DB updates, deployment´s scripts and EXE setups that can be sent to a QA Team to be tested in a new environment.

Deployment

After doing the packaging the next step is the deploy for testing and development use.

For each environment we define an “Environment definition file”. This file allows us to centralize all environment configurations, facilitating consistency. For example, setting all application pool accounts the same, configure port number, define what application will be installed in certain server, etc.

For the deployment step a Job is triggered by Jenkins. This job reads the definition file and triggers others jobs to deploy aech application we need in the environment.

Some jobs are very simple as doing an IISRESET and others more complex as uninstall and install an application with all the configuration needed.

Docker

Next step for us in this improvement is to create a Docker image with Jenkins running inside and with all the jobs we need to create the “final package”. So, then we can create a container for each environment easily.

Also, many of our applications are been dockerized. The containers bring new capabilities that pair with micro-services enables a new era of cloud native applications. It gives us more benefits:

• Increase data center density (basically we can pack more applications in the same hardware).
• Better Scalability. Load Balance and Availability.
• Zero downtime deployment.
• Thriving ecosystem of tool for better operability.

If you are starting a project you can implement this from zero, but if you are working in an old project of many years old (my case), this adaptation can take months or years, that´s why the title of this article reference to the song … It’s a long way to the top …

Would you like to know more? Do you need our help? Contact Us!
www.quadiontech.com

It’s a Long Way to the DevOps was originally published in Quadion Technologies on Medium, where people are continuing the conversation by highlighting and responding to this story.