The Journey of DevOps Continues Promising New Advantages

Authored by Gautam Khanna, Vice President and Head, Modernization Practice, Infosys

Three decades back, in the field of software development, developers were solely responsible for code or design changes to systems including testing and deployment. This ensured quick closure as dependencies were minimal. Then came the era of specialization. Several roles and functions evolved within the organization that included developers, testers, release managers, and many others to optimize effort when scaling to build large systems. However, it also resulted in silos in teams, slowing down the movement of code to the production servers due to multiple handovers and quality issues.

Around 2005, to overcome some of these challenges, several tools evolved to automate the effort required in manual tasks such as reviewing the code against defined standards or test automation. While the tools were successful to an extent, there was a need to bring discipline in their usage. The idea of using more tools and forming unified teams became popular resulting in DevOps getting recognized as a set of practices that combined software development and IT operations in 2008.

With the 2008 downturn in the market, competition became cutthroat. Business needed systems which could quickly publish updates for their end-users. They were looking to constantly innovate on their products and services to attract and retain consumers. Agile methodology was already gaining popularity for faster code development. This need for speed by businesses highlighted two shortcomings in software development. Firstly, there were numerous handover issues amongst these specialized functions including manual errors that added to the cost and schedule overruns. There was a loss of business logic in translation due to a lack of cohesion between the different roles such as business analysts, developers, testers, deployment teams. Secondly, tools lacked all required capabilities or were not being used properly for optimal results in terms of improving quality, reliability, and speed. All these factors led to a real uptake in DevOps in the software industry.

The early takers of DevOps were primarily startups and product companies who wanted speed without compromising quality. Product companies needed multiple and faster releases and DevOps suited them well. Many retailers and telecom companies followed suit, who adopted DevOps in their digital online sales portfolios to reach the market faster. Other portfolios lagged in adoption due to legacy applications and domain products. However, by 2013, these industries saw wider adoption of DevOps across enterprises and several other sectors such as finance and manufacturing companies adopted DevOps in their digital portfolios.

DevOps – Why it matters?

DevOps automation meant using tools for nearly every task in the software lifecycle, except probably code development and test script writing. Developing a software product involves several tasks done by varied teams such as code analysis, unit testing by developers, functional, performance, and other types of testing by testers, stage-wise deployment, and release management. DevOps meant automation of these tasks by using tools in an integrated manner with no manual intervention and automated quality checks between the stages.

Organizations soon realized there was more to DevOps than just achieving early time to market. Investing in DevOps made sense because it saved them a lot of cost and effort. Highly reliable systems meant end-users were happy and stayed loyal. Frequent releases ensured that businesses could attract end consumers and in turn increase their sales. The benefits of DevOps are no longer limited to IT automation but impact business growth and revenue.

Today, all industries are adopting DevOps. Organizations are expanding their DevOps adoption to include non-digital portfolios that support their digital applications to ensure speeding up of the entire value stream. We call this the NextGen DevOps at Infosys. This phase of DevOps adoption is for legacy systems, Mainframes, domain products, ERP, CRM packages, and more. Further adding automated security checks in the DevOps automation is helping organizations avoid million-dollar losses due to frauds.

Not all is achieved in DevOps yet. A strong wave of migrating applications to the cloud and cloud-native development is picking up across industries and DevOps on cloud is an upcoming focus area. Clients are looking at cloud-specific or cloud agnostic DevOps solutions. DevOps for SAAS, AI-ML code, NetOps are also becoming popular. Usage of AI in DevOps automation and continuous improvement in DevOps tools by product vendors promises the emergence of intelligent and efficient tools ensuring DevOps is here to stay.