[bitnami/nginx-ingress-controller] nginx-ingress-controller pod can not be ready, Readiness probe failed

[oscarxie]

Name and Version

bitnami/nginx-ingress-controller 11.6.13

What architecture are you using?

amd64

What steps will reproduce the bug?

1.helm install nginx-ingress-controller
helm install my-release oci://registry-1.docker.io/bitnamicharts/nginx-ingress-controller

2.nginx-ingress-controller always in running, can not be ready

**NAME                                                              READY   STATUS    RESTARTS   AGE**
my-release-nginx-ingress-controller-579666b4f4-2g2r8              **0/1     Running**   0          80s
my-release-nginx-ingress-controller-default-backend-ffb67csbhgz   1/1     Running   0          80s

3.check the pod, found "Readiness probe failed"

  Normal   Pulling    65s   kubelet                   Pulling image "docker.io/bitnami/nginx-ingress-controller:1.12.1-debian-12-r1"
  Normal   Pulled     51s   kubelet                   Successfully pulled image "docker.io/bitnami/nginx-ingress-controller:1.12.1-debian-12-r1" in 13.772s (13.772s including waiting). Image size: 101597766 bytes.
  Normal   Created    51s   kubelet                   Created container prepare-nginx-folder
  Normal   Started    51s   kubelet                   Started container prepare-nginx-folder
  Normal   Pulled     49s   kubelet                   Container image "docker.io/bitnami/nginx-ingress-controller:1.12.1-debian-12-r1" already present on machine
  Normal   Created    49s   kubelet                   Created container controller
  Normal   Started    49s   kubelet                   Started container controller
  Normal   RELOAD     45s   nginx-ingress-controller  NGINX reload triggered due to a change in configuration
  Warning  Unhealthy  36s   kubelet                   Readiness probe failed: Get "http://10.42.0.21:10254/healthz": read tcp 10.236.132.249:40818->10.42.0.21:10254: read: connection reset by peer
  Warning  Unhealthy  26s   kubelet                   Readiness probe failed: Get "http://10.42.0.21:10254/healthz": read tcp 10.236.132.249:56954->10.42.0.21:10254: read: connection reset by peer
  Warning  Unhealthy  16s   kubelet                   Readiness probe failed: Get "http://10.42.0.21:10254/healthz": read tcp 10.236.132.249:39052->10.42.0.21:10254: read: connection reset by peer
  Warning  Unhealthy  6s    kubelet                   Readiness probe failed: Get "http://10.42.0.21:10254/healthz": read tcp 10.236.132.249:43286->10.42.0.21:10254: read: connection reset by peer
  Warning  Unhealthy  6s    kubelet                   Readiness probe failed: Get "http://10.42.0.21:10254/healthz": read tcp 10.236.132.249:43296->10.42.0.21:10254: read: connection reset by peer

Are you using any custom parameters or values?

No

What is the expected behavior?

nginx-ingress-controller pod should be ready

What do you see instead?

I checked https://github.com/bitnami/charts/blob/main/bitnami/nginx-ingress-controller/templates/controller-deployment.yaml, livenessProbe and startupProbe changed to tcpSocket, but readinessProbe use httpGet

          {{- if .Values.customLivenessProbe }}
          livenessProbe: {{- include "common.tplvalues.render" (dict "value" .Values.customLivenessProbe "context" $) | nindent 12 }}
          {{- else if .Values.livenessProbe.enabled }}
          livenessProbe: {{- include "common.tplvalues.render" (dict "value" (omit .Values.livenessProbe "enabled") "context" $) | nindent 12 }}
            tcpSocket:
              port: metrics
          {{- end }}
          {{- if .Values.customReadinessProbe }}
          readinessProbe: {{- include "common.tplvalues.render" (dict "value" .Values.customReadinessProbe "context" $) | nindent 12 }}
          {{- else if .Values.readinessProbe.enabled }}
          readinessProbe: {{- include "common.tplvalues.render" (dict "value" (omit .Values.readinessProbe "enabled") "context" $) | nindent 12 }}
            httpGet:
              path: /healthz
              port: metrics
              scheme: HTTP
          {{- end }}
          {{- if .Values.customStartupProbe }}
          startupProbe: {{- include "common.tplvalues.render" (dict "value" .Values.customStartupProbe "context" $) | nindent 12 }}
          {{- else if .Values.startupProbe.enabled }}
          startupProbe: {{- include "common.tplvalues.render" (dict "value" (omit .Values.startupProbe "enabled") "context" $) | nindent 12 }}
            tcpSocket:
              port: metrics
          {{- end }}

Additional information

No response

[javsalgar]

Hi!

Do the logs show anything meaningful? I cannot reproduce the issue

[oscarxie]

For readinessProbe, the port I change from metrics to http, it works.

    readinessProbe:
      failureThreshold: 3
      httpGet:
        path: /healthz
        **port: http**
        scheme: HTTP
      initialDelaySeconds: 10
      periodSeconds: 10
      successThreshold: 1
      timeoutSeconds: 1

[acfmarcelo]

@oscarxie can you noticed if you got this error too?

(combined from similar events): Error reloading NGINX: exit status 1 2025/04/01 20:26:03 [notice] 1060#1060: signal process started 2025/04/01 20:26:03 [alert] 1060#1060: kill(29, 1) failed (3: No such process) nginx: [alert] kill(29, 1) failed (3: No such process)

After some time, the nginx-ingress-controller Pods crash and not be ready.