Is DevOps The Key Differentiator?

Have you ever wondered what differentiates Google, Amazon, Netflix or any other Software Giant compared to others who wanna be. Is it their super efficient algorithms or use of best possible data structures or some brilliant new product ideas or the use of cutting edge technologies combined with latest research. Of course, importance of these can't be ignored and they do support either in performing optimally or being prepared for future but none of them helps in providing for the user who needs to be served right now.

So what differentiates them is undoubtably, univocally a DevOps culture through out the organisation, DevOps processes supporting the software development & business equally and DevOps tools enabling rapid and reliable software development.

While most of Product Owners try to focus most on new features and enhancements to their existing product they tend to ignore the importance of introducing a DevOps Culture, DevOps Processes and DevOps Tools. Of these the most difficult to introduce is DevOps Culture as it has a human element to it and humans inherently resist change. DevOps processes are easy to integrate once the intention is built and DevOps tools are the easiest to adopt and only require skills to be acquired.

Focusing on collaboration, automation, and agility can have significant benefits, including:

• Faster time to market
• Higher ROI
• Greater user/customer satisfaction
• Increased efficiency of developers and operations
• Improved collaboration
• Early detection and correction of issues

With the premise being set, let me go through each aspect of DevOps at a time for a layman to understand Culture, Process & Tools.

DevOps Culture

Many teams claim to have DevOps culture when they seem to be planning using Agile methods, releasing code with Continuous Delivery or having Automation Testing, although these practices are certainly important and strengthen DevOps but DevOps culture is all about a shared understanding between developers and operations, and sharing responsibility for the software they build. That means increasing transparency, communication, and collaboration across development, IT/operations, and "the business".

It is a big change and can be scary so two ingredients are key: clarity around expectations and an environment of psychological safety. Without them, teams tend to thrash around making mistakes – and, worse, not learn from them.

Four important aspects to be cleared out before adapting DevOps culture can be

1. Roles and Responsibilities - Define who does what, find out any skill gaps and fill them.
2. Rules of Engagement - Define how you'll work together, single team or cross-functional team spanning multiple aspects
3. Pre-Mortem - Anticipating risks so you can solve for them while there's still time! It is far better than doing a Root Cause Analysis
4. Retrospectives - Provide a safe space to discuss what's working, what's not, and what needs to change. These are extremely critical for creating a culture of continuous improvement and learning

DevOps Processes

The DevOps process flow is all about agility and automation. Each phase in the DevOps lifecycle focuses on closing the loop between development and operations and driving production through continuous development, integration, testing, monitoring and feedback, delivery, and deployment.

Continuous Development which is an umbrella term describing iterative process to deliver software to customer. It involves continuous integration, continuous testing, continuous delivery, and continuous deployment.

Most of us are familiar with above image of DevOps Process workflow, let us create a common understanding for future.

Continuous integration

Developers regularly merge code changes into a shared repository where those updates are automatically tested. It helps to prevent delays in development by allowing multiple developers to work on the same source code with confidence, rather than waiting to integrate separate sections of code all at once on release day. It is a crucial component of the DevOps process flow, which aims to combine speed and agility with reliability and security.

Continuous testing

Continuous testing is a verification process that allows developers to ensure the code actually works the way it was intended to in a live environment. Testing can surface bugs and particular aspects of the product that may need fixing or improvement, and can be pushed back to the development stages for continued improvement.

Continuous monitoring and feedback

Throughout the development pipeline, team should have measures in place for continuous monitoring and feedback of the products and systems.

This process allows IT operations to identify issues and notify developers in real time. Continuous feedback ensures higher security and system reliability as well as more agile responses when issues do arise. Throughout the development pipeline, team should have measures in place for continuous monitoring and feedback of the products and systems.

Continuous Delivery

Continuous delivery is the next logical step from CI. Code changes are automatically built, tested, and packaged for release into production. The goal is to release updates to the users rapidly and sustainably.

CD automates the release process (building on the automated testing in CI) so that new builds can be released at the click of a button.

Continuous Deployment

This is the nirvana moment for a team and continuous deployment may be the better option over Continuous Delivery. Continuous deployment is the fully automated version of CD with no human intervention necessary (buttons replaced with probably some bots).

In this process, every validated change is automatically released to users. This process eliminates the need for scheduled release days and accelerates the feedback loop. Smaller, more frequent releases allow developers to get user feedback quickly and address issues with more agility and accuracy.

Continuous deployment is a great goal for a DevOps team, but it is best applied after the DevOps process has been ironed out. For continuous deployment to work well, organisations need to have a rigorous and reliable automated testing environment. If the team is not there yet, starting with Continuous Integration and Continuous Delivery will help get there.  

Googles/Amazons/Netflixes are the ones who have reached this state in their development process and a developer is able to release their code without any human interaction or action.

DevOps Tools

While DevOps is as much about mindset as it is about process, there are still many moving parts to keep track. As team adapt processes to a DevOps environment, consider using DevOps tools to stay on point and can be termed best as enablers.

It is like that even if some has an intent (culture) to clean their house, knowledge (processes) on how to clean it but still they will need a broom to do the work. These tools are like this broom which convert team's intent and knowledge into action for results.

Though this list can be exhaustive and an entire landscapes exist around it, here are the bare minimum tools required to be adapted.

1. Git - Helps to easily save or revert back to different versions of source code and leverage features like branching, staging areas, and different workflow options.
2. JIRA - Just like GPS for travel, Jira Software acts as the single source of truth for development information across DevOps workflow. Gives complete visibility and traceability of work across teams and can surface information and data from integrated toolchain to the right people at the right time. Useful for any team lead, developer, product manger or tester equally.
3. Jenkins - This is the glue which binds all of it together. It has become the open source standard for managing the dev side of DevOps, from source code management to delivering code to production.
4. Automation Testing - Ubiquitous automation is a goal which isn't easy to achieve and there are numerous automation frameworks available. One aspect which many teams tend to ignore is thinking that if they automate testing of their application, that is the best they can do, they tend to forget the performance and integration aspects.
5. Monitoring & Diagnostics - There is a plethora of these out there and selection depends on what are the components (infrastructure/network/application/logs ) that need to monitored. Grafana/Splunk/Prometheus/ELK are popular.
6. Infrastructure Automation - This field has grown a lot over the years and their multiple noteworthy providers in this field vis-a-vis Ansible/ Chef/ Puppet and lately Cloud Infrastructure automation framework, Terraform.

With this we can summarise that though from the top it looks like that it is the quality of engineering talent and bright product minds that fuel the astonishing revenues of Software Giants but in reality it is the adaptation of Reliability Engineering Practices powered by a strong DevOps Culture, well defined DevOps Processes and extensive use of modern DevOps Tools that make them so nimble and agile that they are able to fulfil user's whims and wishes at a rapid pace.

DevOps isn't any single person's job. It's everyone's job.

— Christophe Capel, Principal Product Manager, JIRA Service Desk