With Agility being high on the priority for businesses worldwide, the majority of businesses are looking to deploy software daily and, in some cases, hourly.
The demand for greater agility will see the adoption of DevOps continue to grow in 2019, with many believing that this is the year that it will reach its’ peak popularity and become mainstream.
DevOps popularity to peak in 2019
According to Statista, there has been a 7% increase in the number of businesses that have taken the next step in their DevOps journey going from practicing DevOps in a majority of independent teams to fully embracing DevOps across the enterprise.
With many more businesses having already embraced enterprise-wide adoption of DevOps, the focus in 2019 will be maturing their DevOps practice. The areas of focus will be:
1. Secure DevOps will become a priority with DevSecOps
Security will become a priority; as such businesses will be looking at how they can address the shortcomings of the traditional security model. Most of the current DevOps implementations have focused on ‘shifting left’, moving functional testing earlier into the development process without changing their approach to security. Resulting in security considerations only being addressed at the end of the development cycle.
DevSecOps aims to solve this by treating security as code and integrating security considerations and team members at the start of the development process, making security the entire teams’ responsibility.
2. Automation will remain a priority
Businesses will be looking to reduce cycle time and improve code quality by continuing to invest in test automation, providing better analytics over testing efforts and allowing testing teams to focus on Test Case design rather than execution.
Test automation will continue to be predominately used for regression tests requiring repetitive actions with the focus being extended to cover non-functional requirements, specifically security.
3. Maturing the pipeline with DevOps assembly lines
Businesses will focus on maturing their DevOps pipeline, moving from Continuous Integration (CI) to Continuous Delivery (CD) by moving away from CI pipelines and embracing DevOps assembly lines.
While CI pipelines can be complex workflows, they are developer focused activities for automating build, unit testing and deployments to non-production environments.
DevOps assembly lines glue together various DevOps activities into streamlined, event-driven, end-to-end workflows across teams and tools to allow businesses to visualise and automate their development process from beginning to production.
4. Containerisation and microservices architectures – Kubernetes
Public Clouds have removed the traditional bottlenecks by commoditising infrastructure and platform services paving the way for new application architectures such as the microservices and container-based, and the serverless architecture.
The scalability and efficiency of the microservices and container-based architecture has seen it emerge as the new approach for building distributed mission-critical applications. This architecture allows businesses to use cases to be isolated into small reusable services, eliminating dependencies between systems.
Container orchestration is all about managing the lifecycles of containers and is a critical component when it comes to managing applications, leveraging the microservices and container-based architecture. According to Statista Kubernetes is the most widely used container orchestration tool used by businesses.
5. Get your data ready for AI and ML
DevOps is all about automating and monitoring which involves taking massive amounts of data from different sources; application logs, server logs, build agents, etc, reviewing the data and identify areas for improvement, which is what makes it ideal for AI and ML.
Some of the ways that AI and ML can enrich DevOps are;
• recommend application performance improvements
• estimating the development time of new features/user stories
• improving communications during the development process (chatbots)
• use test data to determining code quality
• reduce cycle-time
The first step is to get your data ready for AI and ML.
If you can think of anything else that will shape the DevOps landscape in 2019 please leave a comment or drop me an email: email@example.com