Moderna plataforma de serviços bancários da FIS
Faça seu banco evoluir com uma plataforma central moderna.
Ongoing Software Development, Continuous Rollout and No Down Time? Say Hello to DevOps. by Barbara Morgan
October 25, 2018
To compete with increasing demands around speed and reliability, many software developers have moved toward new DevOps techniques of continuous development, testing and delivery. This translates to improved performance for businesses. But it’s not without challenges. To fully exploit the opportunities that short development cycles present, companies must understand the process and the technology that supports it.
Shifting Your Culture is the First Step.
Software development used to be a simple but lengthy process, traditionally known as a “waterfall”. Requirements were gathered by analysts, developers wrote code, the testing team tested it, developers corrected it and it finally was released by the implementation team. There were no automation tools, and each group worked autonomously at each stage in the process. In the end, the waterfall could take months or more.
Such large-scale, “big bang” releases are no longer acceptable. Change happens too fast, and companies can’t afford downtime, even if it’s just a few hours.
Ongoing Integration, Delivery and Deployment
In today’s market, agile software teams need rapid feedback to respond to market demands and eliminate problems quickly. Continuous deployment is a software development practice in which every code change goes through the pipeline (integration, delivery, testing and deployment) before it is automatically put into production.
To achieve this, new DevOps methodologies break down the different teams and remove step-by-step processes from the development lifecycle. They also automate testing throughout development. The result is instant feedback, with automated test scripts verifying every change along the way.
Containerization Through the Cloud
Another important DevOps aspect is the adoption of “containerization,” meaning the ability to pack multiple applications into a single software container. This reduces implementation time, while also assuring quality because the software for delivery and deployment is given the availability of self-contained software for rapid delivery and deployment. Container-based software environments can be quickly replicated, template-ized and prepared for production deployments so clients can reap the benefits of their software much earlier.
The use of containerization also sets up easy cloud adoption for reaping benefits such as fault tolerance across multiple region deployment teams, scaling resources up or down, and competitive costs by better infrastructure utilization. Adopting cloud deployment means big improvements in scalability, cost-benefit and fault tolerance across multiple regions.
Automated Test-driven Improvements
By automating the development and testing cycles with the adoption of frameworks and tools, the quality of the software developed increases as the inherent code quality gets better due to multiple quality gates before the code is propagated to live environments. This leads to increased application reliability, availability, resilience characteristics and continuous improvement.
Software releases are no longer spaced far apart. In fact, multiple micro-releases can be made daily to better respond to client and market needs. The use of multiple small software updates released in hours or days means that there are no last-minute additions to requirements as was common when releases were months apart. Market requirements change quickly so these can be accommodated through changes more quickly and effectively, isolated from impact on other areas.
Always-on and Always Improving
With shorter development cycles and the availability of containerized software that can be shared across the enterprise, the groundwork is set for rapid delivery and continuous innovation. Automated test cases make everything predictable and repeatable generating more confidence with clients. With collaborative and feedback-based development, clients can see new products working earlier. Continuous deployment leads to faster innovation, reduced speed to market, customer-centricity, and overall better software quality.