My main use case for GitLab is as a version control system that we are using. Currently, I am working on an end-to-end AI pipeline, and I have deployed my whole code using GitLab so that all things are utilized for version control for my back-end AI and front-end team. We merge all the Git codes into GitLab, and my CI/CD, issue tracking, security, and monitoring is maintained inside GitLab.

We try to collaborate with all the teams together on different features. Within GitLab, we are not utilizing Jenkins or any product management tools because it perfectly renders all the information. It maintains the version control system, which is very helpful to containerize and deploy all services, allowing us to have everything together in production.

CI/CD is the most important feature that I am utilizing with version control and security as well, and all things are very useful inside GitLab.

CI/CD helps my workflow by allowing me to integrate any new changes or any new version that I want to deploy in the whole ML lifecycle, which I implement through new integration phases and identify updates in the deployment scripts. We generate YAML manifest files, add dependencies, and deploy them utilizing GitLab's versioning system, identifying any security patches that need to be added or incidents that need to be managed, triggering the workflow. We try to manage the perfect scenarios.

GitLab has impacted my organization positively in terms of version control systems, providing many smart features and reducing the sharing of dependencies compared to what we used to do previously. It has helped the organization merge and collaborate within the team on the level of code accesses and identify how actionable insights can be inputted within the whole pipelining mechanisms, allowing us to easily perform actions on CI/CD. My organization has adapted this and resulted in more productive work.

There are many improvements that GitLab can implement, such as addressing the issue of caching. Currently, when I have multiple tasks to merge or attempt multiple merges, the CI/CD and overall GitLab processes get slower. Implementing caching to allow parallel jobs to execute together would optimize the solution and enhance efficiency.

The UI is not user-friendly compared to how GitHub Actions operates. If we could customize the UI interface or have options for plugin-based mechanisms, that would be more suitable and increase DevOps in enterprises.

I have been using GitLab for around three years.

GitLab is stable.

GitLab's scalability is good, allowing multiple employees to work together, change code collectively, and perform all kinds of CI/CD regardless of impacts from bug fixes.

I have not reached out to customer support currently, but email support is very good, although I have not interacted with any call-based mechanisms or voice call systems.

I was using GitHub and GitHub Actions prior to this. While GitHub Actions provided good CI/CD operations, GitLab has added smart features, such as a graphical view of branches in version control mechanisms and tracking changes made by authors at specified times. Everything is maintained perfectly in my codebase, and I can easily track buggy code, identifying issues with excellent visibility. This visibility compared to GitHub and GitHub Actions is what led me to switch to GitLab.

The setup cost for GitLab is minimal since the team has its own minimal resource balancing. The costing falls into an intermediate stage and is impactful across all results within the team. It allows for CI/CD stages and addition of security patches smoothly, with only a slight charge that is not significant. Everything related to scripting, processing, management, and deployment works fine.

I have seen a return on investment with increased collaboration within the team resulting in more productive work and a reduction in time based on prior experiences, which emphasizes GitLab's usefulness.

The setup cost for GitLab is minimal since the team has its own minimal resource balancing. The costing falls into an intermediate stage and is impactful across all results within the team. It allows for CI/CD stages and addition of security patches smoothly, with only a slight charge that is not significant. Everything related to scripting, processing, management, and deployment works fine.

Before choosing GitLab, I evaluated options like GitLens and Bitbucket. I have tried those tools prior to this project, but ultimately, I chose GitLab as the foremost solution for version control mechanisms.

Since using GitLab, we have experienced fewer errors in identification, and any incidents coming out at the production level can be maintained to facilitate complete results, ensuring that collaboration works fine and deployment scripts remain easily executable, maintaining all services perfectly. Any impactful scripts work fine, whether making minor or major version updates.

I recommend GitLab if you are looking for a good graph-based solution or any impeccable solution for version mechanisms. I would rate this product a 9 out of 10.

I use ClickHouse and get logs from ClickHouse. I use Loki sometimes. Prometheus is always a thing with Grafana that I use hand-to-hand. Mostly it's Loki and ClickHouse.

I had tried to use TimescaleDB once, but it wasn't an easier option to integrate and set things up, so I just lifted it off and I'm using Loki and ClickHouse for now.

My company doesn't allow me to use any cloud-based services as of now, so everything has to be on-premises and self-hosted. For that reason, I don't use any additional third-party services. But in the near future, when I have the flexibility and liberty of doing it in my organization, I would definitely choose to go for third-party options.

I use GitLab. I use GitLab on our own independent storage servers over here and everything is in-house as I said. We use GitLab and version control all of our codebases and model artifacts and other assets locally in our own storage servers.

The main reason we use GitLab is because we are not supposed to store our code on any other external cloud services. We need to have our own in-house storage solution and in-house version controlling system. Any architecture that we design, we need to have it in-house and air-gapped from the rest of the world. Everything has to be in our control. The bare-metal vertical integration of having servers in our own premises is the reason why I use GitLab.

I appreciate the UI. It's very simple as well as intuitive at the same time. The biggest functionality that I appreciate is that GitLab gives you the liberty to self-host things on your own and keeps things on-premises, as well as being very API-friendly, similar to the original GitHub. There is no cumbersome learning curve to adapt GitLab separately from GitHub. It's a seamless transition switching from the standard GitHub to GitLab.

It's a good UI to work with. GitLab is a consolidation of all of the services. I can use my Docker container builds and my images and store them locally. I can pull those builds and use Docker as well as integrate Docker in the version control pipeline. It becomes very seamless. Sometimes my containers go down or I need to spin up a new node of a computer on its own in my local infrastructure. The system already has a cron job which goes and fetches the appropriate Docker, the latest Docker build with the correct Docker image and it pulls it off.

I would appreciate some more tutorials being put out in the mainstream media like on YouTube, where I could go and learn more about GitLab. Reading a deep dive into the documentation and meddling things on my own, then going and educating my whole team on that is a cumbersome task. I would appreciate if nice, customized YouTube tutorials would be available on YouTube by the official GitLab or maybe by some third-party YouTubers that you guys could partner with, similar to what PyTorch or Weights & Biases have done to democratize the use of their software tools. That would be great on my end.

I have been using GitLab Premium for exactly two years.

I would give a rating of eleven in that case, because we never had a downtime with GitLab Premium. It has never let us down.

I couldn't have an opinion on that because, as I said, we do things in-house. Scalability and the type of auto-scaling or scale-up that we need to do is on my system administrators.

I absolutely am still working with Grafana. For the most part, only partially for some small, tiny projects, I have moved to SigNoz. But for my main, primarily mega, mega foundational projects, the spectacular, the hero projects of my organization, I always prefer Grafana.

It is complex. It takes a good amount of time. Had I had to do it again, I would definitely use an AI agent to do it, because back then two years ago, when I was setting up GitLab Premium on my own in my enterprise, I had to take a good amount of four days to set things up and to seamlessly test everything so that things are working perfectly across my teams.

As of now, I haven't used that. My organization is pretty slim. Our technical team picks up things on their own and does things on their own independently. It's a one-man army kind of developer space in our organization. For that reason, we don't have any cross-communication with developers. Whatever happens, happens verbally, and for that reason, we don't use Jira or any other ticketing solutions.

It's not my responsibility. I have a system admin who looks into the security part of it, but we haven't meddled with any specific functionalities in the security side of it. Whatever we get standard in-house with the default configurations, we just use it.

There are several alternate solutions available.

We don't work in code review. Our developers write their own test cases. Once the test cases pass, the code goes into the specific another department. A solutions architect or a tech lead would be a suitable role for implementation of GitLab Premium. I give this product a rating of ten out of ten.