0

I'm trying to install gitlab using helm on my kubernetes cluster running on vbox.

I have already installed traeffik, metallb and argocd.

I have tried to install gitlab using helm with this command:

helm upgrade \
  --install gitlab gitlab/gitlab\
  --timeout 600s \
  --set [email protected]  \
  --set postgresql.image.tag=13.6.0 \
  -n gitlab

but I have several pods in CrashloopBack due to PVC, PV and others issues.

I have read the documentation, but it is so big that I am lost.

So my question is:

Can I use helm chart to install gitlab on my baremetal cluster or it's only for designed to run on Cloud provider ?

2
  • It would probably help to have more information about why things are failing -- why are pods ending up in CrashLoopBackoff state? What sort of problems are you having with PVCs? It seems likely that you should be able to deploy Gitlab onto a bare metal Kubernetes cluster, but you would need to ensure that you have things like dynamic volume provisioning enabled to support PVCs, etc. Give us more information about why things are failing and we can probably provide more useful answers.
    – larsks
    Commented Jan 4, 2023 at 23:18
  • Oh ok, I did not want to post it in this first thread? Should I post it here or create new post ? Commented Jan 5, 2023 at 11:42

2 Answers 2

1

You probably do not have a default StorageClass.
I had similar problem.
Here is a piece of my own code in ansible provisioning nfs dynamic provider.

  - name: helm install nfs provisioner
    kubernetes.core.helm:
      atomic: true
      chart_ref: nfs-subdir-external-provisioner
      chart_repo_url: https://kubernetes-sigs.github.io/nfs-subdir-external-provisioner
      create_namespace: true
      name: nfs-provisioner
      namespace: nfs-provisioner
      state: present
      wait: true
      values:
        nfs:
          server: 10.1.1.20 (replace with your own)
          path: /srv/nfs/kubernetes/development
        storageClass:
          name: default-nfs
          defaultClass: true
          archiveOnDelete: false

You need to have nfs server installed on some server (in my case it is the same machine)

0

You can use Helm to install GitLab on a bare-metal Kubernetes cluster, as well as on a cloud-based Kubernetes cluster. Helm is a package manager for Kubernetes that makes it easy to install and manage applications on a Kubernetes cluster.

To install GitLab on a bare-metal cluster using Helm, you will first need to install Helm on your cluster. Then, you can use the GitLab Helm chart to deploy GitLab to your cluster. The Helm chart includes all of the necessary resources and configurations to run GitLab on Kubernetes.

To install GitLab using the Helm chart, you will need to run the helm install command and specify the GitLab chart repository, as well as any additional configuration options that you want to set. For example:

    helm install gitlab gitlab/gitlab \
      --set global.hosts.domain=example.com \
      --set [email protected]

This will install the GitLab chart and deploy all of the necessary resources to your cluster. Once the installation is complete, you will be able to access GitLab from your cluster's external URL.

Keep in mind that running GitLab on a bare-metal cluster may require additional configuration and resources compared to running it on a cloud-based cluster. You will need to ensure that you have sufficient resources and network connectivity to support the GitLab installation and any additional workloads that you plan to run on your cluster.

1
  • Hello Dear, Yes I know about helm how it's work because I already used it to install argocd. I did want to know if someone have already installed gitlab on baremetal cluster using helm chart. On official doc it's say " Install GitLab on Kubernetes by using the cloud native GitLab Helm chart." that's why I have doubt Commented Jan 5, 2023 at 11:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.