Search

DevOps: Succeed with the right tools


To set the right expectations about tools, culture is more important in DevOps than tools. For a successful DevOps program you need to take more a holistic CAMS (Culture, Automation, Measure and Share) approach as you read in my previous article on DevOps. Also, before you decide on DevOps tools, building a business case might help you in securing the budget for the tools. While there is no simple way to build a business case for DevOps due to the complex and multi-faceted nature of these programs, the 2016 DevOps report from Puppet Labs is a good start to understand the financial reasons to adopt DevOps. You may specifically like to have a glance at, Chapter 7 - Understanding the ROI of DevOp, within this report.

Coming back to the topic of this article, let’s go through the tools used for DevOps as per various phase of the SDLC:-

Build

Version Control

Historically, centralised version control systems like CVS, SubVersion etc have been the standard for version control. The most obvious drawback with these systems is the single point of failure that the central servers represents. Whenever you have the entire history of the project in a single place, you risk losing everything. This is the reason distributed version control systems (DVCS) like Git, Mercurial etc have become very popular now days. Clients don’t just check out the latest snapshot of the files, they fully mirror the repository. Thus if any server dies, any of the client repositories can be copied back up to the server to restore it. These systems are typically used with commercial hosts like Github or Bitbucket and helps with pull requests, branching strategies, permissions and scalability.

Git, which is the most popular DVCS has a sophisticated branching model. Git allows multiple local branches that can be entirely independent of each other. The creation, merging, and deletion of these branches is very fast. With Git, nearly all operations are performed locally making is blazing fast. The data model used by Git ensures the cryptographic integrity of every bit of your project. Lastly, it’s completely free and open source.

What really sets Git apart are feature branches promoting collaborative coding. The fact that each feature is on a separate branch make it much easier to do code review and generally to collaborate on code. Rather than waiting for change approvals before deploying to production, you can improve code quality and throughput with peer reviews done via pull requests. Pull requests tell your team about changes you’ve pushed to a development branch in your repository. The code reviewer can then review the proposed changes and discuss modifications before integrating them into the main code line.

Integrate

Continuous Integration

Continuous integration is the practice of checking in code to a shared repository frequently and running automated tests each time. This allows you to detect problems early, fix them when they’re cheapest to fix. With branch-and-merge workflows getting popular by using tools like Git, tools that take the pain out of running CI in a multi-branch environment help you to maintain the testing rigor without sacrificing dev speed. You should look for tools that automatically run your tests for secondary branches, and give you the option to push the code to master branch when secondary branch builds are successful. The most popular tool for CI is Jenkins though Bamboo is more popular for environments where JIRA is used due its tight integration with JIRA.

Automated Testing

Unlike manual tests, automated tests are executed faithfully and with the same rigor every time. They also help in identifying the risk areas by producing reports and trend graphs. Automated testing is especially useful in an agile environment where the amount of regression testing exponentially increases as you move to later sprints. It really helps to speed up your development and testing cycles for long running agile projects. Depending on the type of testing various tools can be used e.g. JUnit and NUnit for unit tests, Selenium for GUI testing, SoapUI for API testing, Jmeter and Loadrunner for performance testing, Cucumber and JBehave for behavior driven testing, Jenkins and Bamboo for test orchestration etc. On a different note test driven development inverts the SDLC by starting the development lifecycle with test specs at each stage. Please refer to my article for a more details on agile testing techniques.

Deploy

Containers

Besides the actual work of developing your application to respond appropriately in each environment, you may also face issues with tracking down dependencies, scaling your application, and updating individual components without affecting the entire application. Containers attempt to solve many of these problems. Containers are a way to virtualize operating system that allow you to run an application and its dependencies in resource-isolated processes. Containers allow you to easily package an application's code, configurations, and dependencies into units that deliver environmental consistency, developer productivity, operational efficiency, and version control. Containers help ensure that applications deploy quickly, reliably, and consistently regardless of deployment environment. Docker is the most popular containerization software in use today. These are some of the main benefits of Docker containerization:-

· Isolated lightweight execution environment