DevOps Toolchain with Atlassian – Part 2: Bitbucket

An essential component of any DevOps toolchain is a code repository for all types of software source code, as well as configuration and automation code. In this article, we’ll explore the features and workflows that enable DevOps principles to be put into practice using Atlassian Bitbucket as the central version control system (VCS).

Key topics covered in this article include: 

• Collaboration and communication
• Transparency
• Knowledge management
• Continuous development
• Automation
• Shift Left
• Configuration management

Integration between Jira and Bitbucket

Following on from our latest blog post on the DevOps toolchain with Atlassian, we’ll start with Atlassian Jira. Essentially, integrating these tools improves the traceability of requirements and ensures greater transparency throughout the entire development process. This helps to eliminate, for example, any ambiguity regarding requirements. Furthermore, it reduces the number of errors caused by incorrect assignment or missing information.

Each Jira ticket can be linked to the relevant Bitbucket repository, which houses the software’s codebase. Bitbucket is based on the widely used version control system Git. This makes it possible to create a new development branch directly from the ticket for work on a new feature or bug fix. A branch is a fork of the codebase that serves as a separate workspace. Changes can be made in this branch without altering the codebase used for parallel work or the production software. Atlassian integrates the tools particularly well here, so that, for example, as work progresses on the code, our ticket can automatically be changed from ‘Open’ to ‘In Progress’. 

Knowledge sharing and code quality

Another advantage of Bitbucket is its integrated web interface. All the information and functions are presented very clearly and are easily accessible to users. This means you can view the source code via your browser even without any knowledge of Git command-line commands.

Bitbucket can display and search the complete version history of the code in the browser. This makes it easier, for example, to analyse errors or retrieve old code for a rollback. Furthermore, the code can also be compared (diff) and merged.

A frequently underestimated benefit of a powerful, centralised code management system such as Bitbucket lies in knowledge sharing and code quality. The development history of the software can be traced in Bitbucket. This enables reference solutions to be reused or decisions to be understood. This makes it easier for junior software developers to get started and encourages the exchange ofknow-how amongst colleagues right from the start.

Configuration as Code

Bitbucket should not only contain the software code. A high degree of automation is always accompanied by numerous scripts or declarative descriptions. Examples include scripts for deploying the software (Bash, PowerShell, Ansible, Puppet, etc.), infrastructure definitions (Terraform, CloudFormation, ARM, HELM, etc.) or container descriptions (Dockerfiles, etc.). The definition of a so-called pipeline – starting from the commit, through build, test, integration, delivery, release and deployment – should also be described declaratively wherever possible. All this information and these files can also be stored in Bitbucket. The benefits described so far – collaboration, transparency, integration, reviews and knowledgesharing – can thus be utilised immediately.

Bitbucket can be easily extended using apps from the Atlassian Marketplace or integrated with other components of the toolchain via webhooks.

In my next blog post, I’ll be looking at the CI/CD tool Atlassian Bamboo and its integration with Bitbucket. I’ll be examining the build and deploy phases, as well as the CI/CD pipeline in the context of DevOps, in more detail.