Helm Chart - Global configuration in Helm values.yaml creates CrashLoopBackoff

[bruxillensis]

Describe the bug
When using the globalConfiguration option in the Helm configuration file (i.e. values.yaml) in a more or less vanilla deployment of NGINX results in crashloop backoff.

To Reproduce
Steps to reproduce the behavior:

1. helm install cashew nginx-stable/nginx-ingress -f values.yaml -n default
2. kubectl get pods

NAME                                                     READY   STATUS             RESTARTS   AGE
cashew-nginx-ingress-5fc9bf67b-gfzjh                     0/1     CrashLoopBackOff   1          12s

3. kubectl logs <pod-name>

I0820 22:53:47.214446       1 main.go:214] Starting NGINX Ingress controller Version=1.8.1 GitCommit=2c9e223b
F0820 22:53:47.237979       1 main.go:385] Error when getting default/cashew-nginx-ingress: globalconfigurations.k8s.nginx.org "nginx" not found

Content of values.yaml

controller:
    globalConfiguration:
        create: true
        spec:
            listeners:
            - name: redis-tcp
              port: 6379
              protocol: TCP

Expected behavior
Launch and open port 6379 for TransportServers

Your environment

• Version of the Ingress Controller - 1.8.1
• Version of Kubernetes

kubectl version

Client Version: version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.6", GitCommit:"dff82dc0de47299ab66c83c626e08b245ab19037", GitTreeState:"clean", BuildDate:"2020-07-15T16:58:53Z", GoVersion:"go1.13.9", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.6", GitCommit:"dff82dc0de47299ab66c83c626e08b245ab19037", GitTreeState:"clean", BuildDate:"2020-07-15T16:51:04Z", GoVersion:"go1.13.9", Compiler:"gc", Platform:"linux/amd64"}

• Kubernetes platform (e.g. Mini-kube or GCP) - microk8s

microk8s ctr version

Client:
  Version:  v1.2.5
  Revision: bb71b10fd8f58240ca47fbb579b9d1028eea7c84

Server:
  Version:  v1.2.5
  Revision: bb71b10fd8f58240ca47fbb579b9d1028eea7c84

• Using NGINX or NGINX Plus - NGINX

Additional context
I am unsure if this is a template issue or my configuration of k8s.

[bruxillensis]

There is a mistake in the above error log. The real error log is this:

I0820 23:21:10.322653       1 main.go:214] Starting NGINX Ingress controller Version=1.8.1 GitCommit=2c9e223b
F0820 23:21:10.344992       1 main.go:385] Error when getting default/cashew: globalconfigurations.k8s.nginx.org "cashew" not found

The output of kubectl get globalconfigurations is

NAME                   AGE
cashew-nginx-ingress   109s

The NGINX instance seems to be looking for the incorrect resource name.

[pleshakov]

Hi @bpoulse

It's a bug in our helm chart. Thanks for reporting. We will fix it.

In the meantime, you can try the following workarounds:

if you install the helm chart from the sources.

(1) Modify the following file to fix the issue:

diff --git a/deployments/helm-chart/templates/controller-deployment.yaml b/deployments/helm-chart/templates/controller-deployment.yaml
index 4349141f..6b952d25 100644
--- a/deployments/helm-chart/templates/controller-deployment.yaml
+++ b/deployments/helm-chart/templates/controller-deployment.yaml
@@ -141,7 +141,7 @@ spec:
           - -enable-tls-passthrough={{ .Values.controller.enableTLSPassthrough }}
           - -enable-snippets={{ .Values.controller.enableSnippets }}
 {{- if .Values.controller.globalConfiguration.create }}
-          - -global-configuration=$(POD_NAMESPACE)/{{ .Release.Name }}
+          - -global-configuration=$(POD_NAMESPACE)/{{ include "nginx-ingress.name" . }}
 {{- end }}
 {{- end }}
           - -ready-status={{ .Values.controller.readyStatus.enable }}

(2) Install the chart using the modified sources:

$ cd deployments/helm-chart
$ helm install cashew  -f values.yaml -n default .

Or alternatively, deploy a GlobalConfiguration resource without helm with the name cashew

[bruxillensis]

Thanks for the reply and fix! Both options indeed work for the interim.