How to Install GitLab with GitLab Operator on OpenShift 4 Local?

I am trying to setup a local sandbox on my pc to play with OpenShift and GitLab CICD. I found instructions in multiple resources and tried to apply them, but I am running into an issue with the GitLab installation. After I apply the GitLab deployment using the GitLab Operator, I can see the routes, but when I use the webservice route I get an error.

Application is not available
The application is currently not serving requests at this endpoint. It may not have been started or is still starting.

Possible reasons you are seeing this page:

The host doesn't exist. Make sure the hostname was typed correctly and that a route matching this hostname exists.
The host exists, but doesn't have a matching path. Check if the URL path was typed correctly and that the route was created using the desired path.
Route and path matches, but all pods are down. Make sure that the resources exposed by this route (pods, services, deployment configs, etc) have at least one pod running.

I am using OpenShift Local which is installed with a msi on windows 11 pro. The install will add crc and which uses windows’ Hyper-V.

The setup I picked was

crc setup
crc config set memory 16384
crc config set disk-size 100
crc config set cpus 6
crc start

After the OpenShift cluster is up, I begin to fallow directions which I found from Red Hat and GitLab.

Which follows:

Step 1: Prerequisites

Deploy a Custom SCC

allowHostDirVolumePlugin: false
allowHostIPC: false            
allowHostNetwork: false                     
allowHostPID: false                         
allowHostPorts: false  
allowPrivilegeEscalation: true
allowPrivilegedContainer: false
allowedCapabilities: null
apiVersion: security.openshift.io/v1        
defaultAddCapabilities:         
- NET_BIND_SERVICE                              
fsGroup:                        
  type: MustRunAs            
groups: []                   
kind: SecurityContextConstraints
metadata: 
  name: gitlab-nginx-ingress-scc
priority: null   
readOnlyRootFilesystem: false
requiredDropCapabilities:
- ALL            
runAsUser:         
  type: MustRunAs
  uid: 101
seLinuxContext:
  type: MustRunAs

Deploy the IngressClass

apiVersion: networking.k8s.io/v1
kind: IngressClass
metadata:
  name: gitlab-nginx
spec:
  controller: "k8s.io/ingress-nginx"

Deploy cert-manager

Install cert-manager operator in default namespace using OpenShift’s Operator Hub.

Step 2: Install GitLab Operator

Install the GitLab Operator using OpenShift’s Operator Hub.

Create a GitLab Deployment using Red Shift’s Operator Lifecycle Manager (OLM)

Find domain name associated with the Route created for Webservice

kubectl get route -n openshift-console console -ojsonpath='{.status.ingress[0].host}'
console-openshift-console.apps-crc.testing

Modify the yaml to use OpenShift Routes for Ingress

kind: GitLab
apiVersion: apps.gitlab.com/v1beta1
metadata:
  name: gitlab
  namespace: gitlab-system
spec:
  chart:
    values:
      nginx-ingress:
        enabled: false
      global:
        hosts:
          domain: apps-crc.testing
        ingress:
          class: none
          annotations:
            route.openshift.io/termination: "edge"
    version: 8.1.2

Now I check results and see that things are not ok

I see that that the gitlab CRD is stuck as pending for status

oc -n gitlab-system get gitlab
NAME STATUS VERSION
gitlab   Preparing   8.1.2

But all the pods seem to be running, unless there are some that are missing completely

oc -n gitlab-system get pods
NAME                                         READY   STATUS      RESTARTS   AGE
gitlab-controller-manager-bccdfd4bd-xr8vk    2/2     Running     0          16m
gitlab-gitaly-0                              1/1     Running     0          2m3s
gitlab-gitlab-exporter-ff6598bc8-4kqgt       1/1     Running     0          92s
gitlab-gitlab-shell-85c5f78464-6nz4j         1/1     Running     0          27s
gitlab-gitlab-shell-85c5f78464-jn7td         1/1     Running     0          93s
gitlab-kas-5f6fc9556f-4r5hb                  1/1     Running     0          92s
gitlab-kas-5f6fc9556f-l2m2h                  1/1     Running     0          27s
gitlab-migrations-8defcdb-243-1-5vjx7        0/1     Completed   0          92s
gitlab-minio-69bb7f84bb-87r96                1/1     Running     0          2m3s
gitlab-postgresql-0                          2/2     Running     0          2m3s
gitlab-redis-master-0                        2/2     Running     0          2m3s
gitlab-registry-84f4c4f4d4-cvms5             1/1     Running     0          93s
gitlab-registry-84f4c4f4d4-x49gw             1/1     Running     0          27s
gitlab-shared-secrets-0478af5-6bxlp          0/1     Completed   0          2m13s
gitlab-sidekiq-all-in-1-v2-7cc4bb658-krppj   1/1     Running     0          42s
gitlab-toolbox-79ccb76d9-2c74f               1/1     Running     0          92s

I do see the routes now

oc -n gitlab-system get routes
NAME                              HOST/PORT                   PATH          SERVICES                    PORT                   TERMINATION     WILDCARD
gitlab-kas-5pq82                  kas.apps-crc.testing        /k8s-proxy/   gitlab-kas                  tcp-kas-k8s-api        edge/Redirect   None
gitlab-kas-cd6hx                  kas.apps-crc.testing        /             gitlab-kas                  tcp-kas-external-api   edge/Redirect   None
gitlab-minio-88zf6                minio.apps-crc.testing      /             gitlab-minio-svc            http                   edge/Redirect   None
gitlab-registry-5khcg             registry.apps-crc.testing   /             gitlab-registry             http-registry          edge/Redirect   None
gitlab-webservice-default-dm2ld   gitlab.apps-crc.testing     /             gitlab-webservice-default   http-workhorse         edge/Redirect   None

But when I try to access the webserver

https://gitlab.apps-crc.testing

I get an error message:

Application is not available
The application is currently not serving requests at this endpoint. It may not have been started or is still starting.

Possible reasons you are seeing this page:

The host doesn't exist. Make sure the hostname was typed correctly and that a route matching this hostname exists.
The host exists, but doesn't have a matching path. Check if the URL path was typed correctly and that the route was created using the desired path.
Route and path matches, but all pods are down. Make sure that the resources exposed by this route (pods, services, deployment configs, etc) have at least one pod running.

So, that is where I am currently stuck. If I look in the logs of the gitlab-controller-manager- I can see something, but not sure if the is the result of not being able to access the route.

{"level":"info","ts":"2024-07-12T20:21:48Z","logger":"controllers.GitLab","msg":"ensuring Webservice and Sidekiq are reconciled if enabled","gitlab":{"name":"gitlab","namespace":"gitlab-system"}}
{"level":"info","ts":"2024-07-12T20:21:48Z","logger":"controllers.GitLab","msg":"Webservice and/or Sidekiq not yet Running","gitlab":{"name":"gitlab","namespace":"gitlab-system"}}

Any Ideas?

Resources

https://www.redhat.com/en/blog/install-the-gitlab-operator-on-openshift
https://www.youtube.com/watch?v=nJDCUfbEmVI
https://docs.gitlab.com/operator/openshift_ingress.html#openshift-routes

Filed under: Kiến thức lập trình - @ 03:26

Thẻ: windowskubernetesgitlabopenshiftredhat

Thiết kế website giá rẻ

Danh mục

How to Install GitLab with GitLab Operator on OpenShift 4 Local?