GitLab Ultimate Self-Managed

I have been using GitLab  since I started my journey in IT because GitLab  is important for all software developers, DevOps, and SREs in all fields in IT. I have been using it for a long time.

My main use case for GitLab is source code management combined with CI/CD automation. I use GitLab to host application and infrastructure code, manage branches, and merge requests, and run automated CI/CD pipelines that build, test, and deploy applications across every environment. As an SRE-focused DevOps professional, I primarily use GitLab for end-to-end DevOps workflows from version control to automated CI/CD, triggering pipelines on every commit, running tests and security scans, building Docker  images, and deploying to cloud or Kubernetes  environments using GitLab Runners. This is how I use it day-to-day.

Using GitLab for this DevOps workflow has significantly improved my efficiency and my team's efficiency by centralizing the entire DevOps workflow, code, CI/CDs, reviews, and deployments in one platform. This reduces tool hopping and makes collaboration much smoother. First, GitLab enables faster and safer deployments. Standardized pipelines and approval-based merge requests ensure consistent deployments across environments, reducing production issues. Secondly, quick issue detection and rollback are facilitated through pipeline failures and job logs, which help identify problems early, and version releases make rollbacks faster and safer. Automation reduces manual work as CI/CD pipelines automatically build, test, scan, and deploy on every commit, saving hours of manual effort and eliminating human errors. The fourth point is infrastructure as code at scale; managing Terraform  and Ansible  code in GitLab allows repeatable, auditable infrastructure changes with clear history. Finally, improved reliability and confidence arise because automated testing and security scans increase confidence in releases and reduce post-deployment incidents.

Another valuable aspect is better collaboration and visibility, which comes with merge requests, inline reviews, and pipeline status checks, making it easy for the team and me to review changes and catch issues early.

GitLab has had a significant overall positive impact on my organization by standardizing and automating how we build, test, and deploy software. Having code management, CI/CD, security, and collaboration in a single platform improves speed, reliability, and transparency across teams.

The improvements have resulted in faster and more reliable releases. We replaced manual deployments with automated CI/CD pipelines, which have made releases predictable and repeatable, with deployment time reduced from hours to minutes. Another improvement is reduced production incidents; mandatory pipeline tests and approvals before merges and early failure detection through automated checks lead to standardized deployment processes across environments.

Additional improvement comes through shift-left security, where security scanning built directly into pipelines detects vulnerabilities early rather than in production, eliminating the need for separate security tools for basic scanning. Finally, better onboarding and knowledge sharing occur through standardized CI/CD templates and documented pipelines, allowing new team members to become productive faster while reducing dependency on tribal knowledge.

In my view, the best features GitLab offers include integrated CI/CD, which is one of GitLab's strongest capabilities. We define pipelines in a .gitlab-ci.yml file and runners execute them automatically on commits and merge requests. It automates building, testing, and deploying, eliminates manual release steps, and includes quality, security, and compliance stages, while also being easy to scale with GitLab Runners, leading to faster feedback loops, fewer human errors, and consistent deployments.

The second feature I would mention is merge requests, which combine collaboration, review, and automation in one place. This impacts better code quality, cleaner history, and structured team collaboration. The third point is built-in security scanning, as GitLab offers automatic scanning integrated into pipelines including SAST , DAST, and dependency scanning, making security a part of our CI/CD pipeline rather than an afterthought.

Lastly, pipeline visualization and insights help us understand delays or failures with graphical pipeline views, job logs, and metrics such as pipeline duration and failure rates.

A pain point I have encountered with GitLab is that large GitLab-ci.yml files become hard to read and maintain. YAML syntax is strict, and errors are easy to make, while debugging pipeline logic can sometimes take time, leading to slower iteration when the pipeline grows complex. I propose an improvement idea of better visual pipeline editors and stronger validations and linting before a commit.

Regarding runner management and scaling, managing self-hosted runners requires effort and scaling runners during peak usage can be challenging, which leads to pipeline delays during high load. An expected improvement here is smarter auto-scaling by default, along with better runner health visibility and alerts.

Concerning security features, advanced security scans are locked behind paid tiers, limiting coverage for smaller teams unless the budget allows. I suggest introducing more basic security features in the free tier and clearer guidance on prioritizing vulnerabilities.

Another area for improvement is UI performance and navigation. Finding older pipelines or logs or settings sometimes takes extra clicks, leading to small but noticeable productivity loss. An improvement would be a faster UI for large repositories and enhanced global search and filtering.

I started my journey in 2021, and since my first organization, I have been working in my current field as a Site Reliability Engineer for nearly five years.

GitLab is pretty stable in my experience. I have not experienced any downtime or reliability issues so far.

GitLab's scalability is really great, and it handles growth in users, projects, or workloads effectively, helping us in every aspect.

I have not interacted with GitLab's customer support because I have not faced downtime or any significant issues while using GitLab. Therefore, my experience is great since I have not encountered any issues requiring support.

Positive

I have not used any other solution before GitLab, as I started using GitLab right from the beginning of my career in IT.

My experience with GitLab's pricing, setup cost, and licensing is very positive. GitLab follows a tier-based licensing model that includes free, premium, and ultimate options. I have experience with the free tier and later evaluated the paid tiers mainly for advanced security, compliance, and governance features. I found pricing reasonable for the value, especially when compared to buying separate tools for CI/CD, security, and repo management.

I can share concrete ROI metrics and examples that focus on time saved, costs avoided, and risks reduced. For instance, with deployment automation, before GitLab, it took two engineers one to two hours per deployment, but after implementing GitLab CI/CD, it only needs one engineer and takes 10 to 15 minutes, resulting in a 70 to 85% reduction in deployment effort and saving dozens of engineering hours per month. Additionally, we see cost avoidance from reduced production incidents, so automated tests, approvals, and pipelines minimize human error, showing a 30 to 40% reduction in deployment-related incidents, which results in less downtime and fewer after-hours escalations.

After adopting GitLab, I can share some measurable outcomes. Before GitLab, deployment time took one to two hours for manual steps and coordination, and now it is down to 10 to 15 minutes, reflecting a 75 to 85% reduction in deployment time. Regarding release frequency, previously we had one to two releases per week, but now we achieve daily or on-demand releases, resulting in a three to five-fold increase in release frequency. In terms of production incidents, we faced frequent post-deployment issues before GitLab, but we now see a noticeable drop due to automated tests and approvals, with a 30 to 40% reduction in deployment-related incidents.

The setup cost was moderate and not very high. For GitLab SaaS, the initial setup cost was minimal, while self-managed GitLab involved infrastructure, VM storage backups, runner configuration, and integrations, which I also found moderate.

Before choosing GitLab, I did evaluate other options, specifically Bitbucket  in combination with Jenkins . We started exploring Bitbucket , but after weighing the pros and cons, I decided to move to GitLab.

I would advise others looking into using GitLab to definitely go for it because it has really good features. Start simple, then scale by not building complex pipelines from day one; instead, begin with basic build, test, and deploy stages, adding security scans, approvals, and optimizations gradually. This keeps pipelines readable and reduces early frustrations.

Use reusable CI/CD templates to create organization-wide pipeline standards, which improves consistency, decreases duplication, and speeds up onboarding. Additionally, invest early in a runner strategy to decide between SaaS runners and self-hosted runners, planning for auto-scaling and isolation to prevent runner bottlenecks. Finally, monitor and optimize pipelines by tracking duration and failure rates, removing slow or flaky jobs to ensure fast feedback that keeps developers engaged and productive.

The reviews I provided are genuinely positive because I find GitLab to be an excellent product for us in IT. Honestly, there are other products in the market that serve as alternatives to GitLab, but I cannot envision working in IT, especially in servers, deployment, or CI/CD, without GitLab. I am a huge fan of GitLab, and my experience working with it has been wonderful; it has greatly aided our teams by reducing human errors and the number of personnel required. My overall review rating for GitLab is 9 out of 10.

GitLab  serves as my main DevOps platform for managing our repository and code base, as well as for pipeline integration of our system.

We maintain all our projects in GitLab  with separate folders for project files, and we use GitLab to manage our projects. We also use GitLab pipeline to deploy in our dev environment, QA environment, and production, as well as for creating patches.

GitLab has positively impacted my organization by being faster than other platforms and providing the best user interface and features. Creating merge requests is easy, and it makes it simple to use for new joiners.

GitLab's best features are its built-in CI/CD and pipeline integration, which can be easily connected to AWS  or other deployment platforms. The CI/CD pipeline integration is the most valuable aspect for me, and it also provides an easy user interface to create merge requests, merge, and create branches.

The CI/CD pipeline integration has helped my workflow by allowing us to use the .gitlab-ci.yml file to integrate and write our pipeline codes, where we include build jobs, test jobs, and deploy jobs directly with our AWS  for S3  using CI/CD variables for push.

In GitLab pipelines, both manual and automatic integration are offered for the pipeline runner, and pipeline logs are provided, which are very useful for DevOps engineers and developers for debugging.

GitLab can be improved by being more responsive in the UI and offering better pricing for premium features, which would be useful for small startups. While GitLab's CI/CD is powerful, it is somewhat complex in certain aspects.

It would be better if the GitLab and Jira  integration were more flexible and easier to integrate with Jira  for task management in the future.

Since I joined my company, we have been using GitLab for our repo maintenance, so I have been using it for more than two years.

In my experience, GitLab is stable; however, sometimes it takes much time to start the runner and gets stuck in a pending situation, possibly due to traffic issues.

GitLab handles growing workloads and users well, and its scalability is good, but the downtime issue is a concern because it sometimes takes too long to start the runner.

Earlier, I did my personal projects using GitHub  before I joined my current organization, which uses GitLab, and that is the reason for switching. I personally prefer GitLab over GitHub  because it is user-friendly and easy to work with.

My current organization started with GitLab.

I would advise others looking into using GitLab that it is a perfect platform for organization-level repository management and pipeline CI/CD features are very useful and easy to use.

I have shared everything I feel and noted the pros and cons of GitLab as a user. Overall, I rate GitLab positively, giving it a seven out of ten because I have been using GitLab for more than two and a half years and have not faced many difficulties. However, in some cases, I got stuck with runners getting delayed to run, which can be annoying.

Positive

Neutral