Troubleshooting and the Road Ahead with GitLab

By now, we have covered the end-to-end use of GitLab CI/CD for planning, building, testing, and shipping software. You should hopefully be familiar with the vocabulary of GitLab CI/CD pipelines and Runner infrastructure, and you should also have the confidence to develop and deploy basic applications with GitLab.

CI/CD occupies a constantly shifting space in the software industry. The best practices and tools of the trade today may very well be obsolete five years from now, if not earlier. We have made a conscious effort in this book to maintain a balanced focus between concepts and tooling. The goal is to preserve relevance by emphasizing CI/CD and DevOps fundamentals, while still offering you a chance to practice and follow along, even if the syntax and some of the tools change over the next several years.

The purpose of this final chapter is to synthesize what we’ve learned and guide you toward the next steps in your DevOps...

Technical requirements

As in previous chapters, an account on a GitLab instance (SaaS or self-managed) is recommended. Some of the content in this chapter is more conceptual than example-based, but access to Infrastructure-as-Code (IaC) tooling such as Terraform and Ansible is recommended if you wish to practice the concepts discussed in the Managing your operational infrastructure using GitOps section. In that case, an account in a cloud service provider (such as AWS or Microsoft Azure) would be recommended for infrastructure provisioning and configuration management using your chosen IaC tool.

We will now turn to identifying and troubleshooting common issues when working with GitLab CI/CD.

Troubleshooting and best practices for common pipeline problems

Problems encountered in GitLab CI/CD pipelines tend to fall into two broad categories. The first category is errors or unexpected behavior caused by the syntax and logic in .gitlab-ci.yml, the primary CI/CD configuration file. The second category involves limitations or misconfigurations in the runner infrastructure used to run CI/CD jobs. We will address each of these categories in turn.

Troubleshooting CI/CD syntax and logic

When errors in GitLab CI/CD can be narrowed down to the content of .gitlab-ci.yml, the first step in troubleshooting is to identify whether the issue is faulty or unsupported syntax, or misconfiguration (or misunderstanding) of an otherwise valid YAML file.

Syntax errors in .gitlab-ci.yml

One of the easiest cases to troubleshoot (and therefore a good place to check for problems first) is syntax or formatting errors in .gitlab-ci.yml. A typical example is errors caused by omitting required...

Managing your operational infrastructure using GitOps

Development cannot be divorced from the infrastructure executing the code. We’ve seen this clearly as we’ve journeyed from using GitLab as a source code management tool to defining CI/CD pipelines to define how our source code is built and deployed, with runners (infrastructure) as an essential component of CI/CD.

DevOps properly conceptualized is a culture unifying development and operations. How, then, can we incorporate configuring and managing our operational infrastructure into the GitLab flow already used and understood by developers? The answer lies in implementing GitOps. GitOps is a development practice that falls under DevOps, which encourages using a similar iterative change management model already used by development teams. As the term suggests, Git is an essential part of the workflow. Infrastructure should be under version control, just like application source code. Changes to infrastructure are then...

Future industry trends

It goes without saying that learning anything new comes with the expectation that the knowledge will remain relevant in the future. Looking forward, continuous integration and delivery are well positioned to persist as critical practices in the software landscape, while continuing to undergo evolutionary changes. We can identify three likely ways the software development life cycle will change in the coming years. Those trends are increased automation, increased abstraction, and reduced development cycle time. We can also expect that GitLab as a platform will continue to play an important role. The concepts, examples, and tools presented in the previous chapters will serve as a good foundation to be a more productive developer right now while providing the confidence needed to adapt as the industry inevitably shifts.

Automation will create more software at a larger scale

Ask any organization today what is the most expensive part of operating, and they will...

Conclusion and next steps

This book has guided you through using GitLab as a platform to manage your software development life cycle. After discussing the state of the industry before DevOps emerged as a culture and methodology, you were then guided through the basics of Git version control and GitLab’s project management components. Then, GitLab CI/CD pipelines were introduced as the central feature for organizing, designing, and automating your development workflow.

You learned that GitLab CI/CD comprises three components: the text-based pipeline configuration file, the runners that execute pipeline tasks, and the main GitLab application that coordinates between the configuration and the runner agents. We then walked through how to leverage pipelines to verify, secure, package, and deploy your code. You learned that you can use GitLab CI/CD as a unified tool to perform these often-disparate steps in the software life cycle.

After covering the supported CI/CD features...