DevOps principles

DevOps is based on a set of foundational beliefs and processes. These form the pillars on which it is built and provide a natural ecosystem for the delivery of excellence within an organization. Let's look briefly into some of these principles.

Collaboration and communication

One of the prime tenets of DevOps is collaboration. Collaboration means that different teams come together to achieve a common objective. It defines clear roles and responsibilities, overall ownership, accountability, and responsibility for the team. The team comprises both development and operations people. Together they are responsible for delivering rapid high-quality releases to the end customer.

Both teams are part of the end-to-end application life cycle process. The operations team contributes to the planning process for features, providing their feedback on overall operational readiness and issues regarding business application and services. Concurrently, the development team must play a role in operational activities. They must assist in deploying the release to production and provide support in terms of fixing any production issues that arise. This kind of environment and ecosystem fosters continuous feedback and innovation. There is a shared vision, where everyone in the team are working toward common goals.

Flexible to change

Agility refers to the flexibility and adaptability of people, processes, and technology. People should have a mindset open to accepting change, playing different roles, and taking ownership and accountability. Processes would generally refer to the following:

• Application lifecycle management
• Development methodology
• Software design

Application lifecycle management

Wikipedia defines application lifecycle management as follows:

Application lifecycle management (ALM) is the product lifecycle management (governance, development, and maintenance) of computer programs. It encompasses requirements management, software architecture, computer programming, software testing, software maintenance, change management, continuous integration, project management, and release management.

Application lifecycle management (ALM) refers to the management of planning, gathering requirements, building and hosting code, testing code in terms of code coverage, unit tests, versioning of code, releasing code to multiple environments, tracking and reporting, functional tests, environment provisioning, deployment to production, and operations for business applications and services. The operational aspects include monitoring, reporting, and feedback activities. Overall, ALM is a huge area and comprises multiple activities, tools, and processes. Special attention should be given to crafting appropriate application lifecycle steps to induce confidence in the final deployed system. For example, processes can be implemented which mandate that code cannot be checked in the source code repository if unit tests do not pass completely. ALM comprises multiple stages such as planning, development, testing, deployment, and operations.

In short, ALM defines a process to manage an application from conception to delivery and integrates multiple teams together to achieve a common objective. The phases of a typical application lifecycle management process is shown in Figure 1. ALM is a continuous process that starts with the planning of an iteration, building and testing the iteration, deploying it on a production environment, and providing post-deployment services to the customer.

Feedback from customers and operations is passed on to the planning team, which eventually incorporates them into subsequent iterations, and this process loop continues.

Figure 1: Application lifecycle management phases

Development methodology

Development methodology should be flexible and elastic to enable multiple smaller iterations or sprints of delivery. Each sprint and iteration must be functionally tested. Smaller iterations help in completing specific smaller features and pushing them to production. This provides the team with a clear sense of the direction and scope of the work, raising expectations and giving them a sense of ownership over the release.

Software design

Software design should implement architectural principles that foster modularity, decomposition of large functionality into smaller features, reliability, high availability, scalability, audit capabilities, and monitoring, to name a few.

Automating processes and tools

Automation plays an important role in achieving overall DevOps goals. Without automation, DevOps cannot achieve its end objectives. Automation should be implemented for the entire application lifecycle management, from building the application, to delivery and deployment to the production environment. Automation brings trust and a high level of confidence in the output from each phase of the software development life cycle.

The probability that deliverables are of high quality, robust, and relatively risk-free is quite high. Automation also helps in the rapid delivery of a business application to multiple environments because it is capable of running multiple build processes, executing thousands of unit tests, figuring out code coverage comprising millions of lines of code, provisioning environments, deploying applications, and configuring them at the desired level.

Failing fast and early

At first glance, it seems weird to talk about failure in a DevOps book that is supposed to assist with the successful delivery of software. Trust me, it is not! Failing fast and early refers to the process of finding issues and risks as early as possible within application lifecycle development. Not knowing the issues that arise toward the end of the ALM cycle is an expensive affair because a lot of work has already been done on it. Such issues might involve making design and architectural changes, which can jeopardize the viability of the entire release. If the issues can be found at the beginning of the cycle, they can be resolved without much impact to the release. Automation plays a big part in identifying the issues early and fast.

Innovation and continuous learning

DevOps fosters a culture of innovation and continuous learning. There is a constant feedback flow regarding the good and bad, and what's working and what's not working in various environments. The feedback is used to try out different things, either to fix existing issues or find better alternatives. Through this exercise, there is a constant information flow about how to make things better and that in turn provides the impetus to find alternative solutions. Eventually, there are breakthrough findings and innovation, which can be further developed and brought to production.