CODEWITHSUNDEEP
kubernetesamazon-eksmetabase
Gijsriet
Gijsriet

Reputation: 163

New service in existing Kubernetes Cluster gives 503 Temporarily Unavailable

I'm trying to run a Metabase deployment in an existing, working, EKS kubernetes cluster, but keep getting 503 Service Temporarily Unavailable.

The pod is running, the ports seem to be assigned correctly and I have an Application Load balancer running and forwarded to the metabase service.

Service:

apiVersion: v1
kind: Service
metadata:
  name: metabase-service
  namespace: default
spec:
  ports:
  - port: 80
    protocol: TCP
    targetPort: 3000
  selector:
    app: metabase
  type: ClusterIP

Deployment

apiVersion: apps/v1
kind: Deployment
metadata:
  generation: 1
  labels:
    deployment: metabase
  name: metabase
  namespace: default
spec:
  replicas: 1
  selector:
    matchLabels:
      app: metabase
  template:
    metadata:
      labels:
        app: metabase
    spec:
      containers:
      - name: metabase
        image: metabase/metabase
        imagePullPolicy: Always
        ports:
        - containerPort: 3000
          protocol: TCP

Ingress

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: my-ingress
  namespace: default
spec:
  ingressClassName: alb
  rules:
  - host: staging.my-app.nl
    http:
      paths:
      - backend:
          service:
            name: my-app-service
            port:
              number: 80
        path: /
        pathType: Prefix
  - host: metabase.my-app.nl
    http:
      paths:
      - backend:
          service:
            name: metabase-service
            port:
              number: 80
        path: /
        pathType: Prefix
  tls:
  - hosts:
    - staging.my-app.nl
    secretName: myapp-certs-2022
  - hosts:
    - metabase.my-app.nl
    secretName: myapp-certs-2022

output from kubectl get ingress NAME CLASS HOSTS ADDRESS PORTS AGE

my-ingress   alb     staging.my-app.nl,metabase.my-app.nl   k8s-default-my-app-12334444444.eu-central-1.elb.amazonaws.com   80, 443   206d

output from kubectl get svc

metabase       ClusterIP   10.100.127.227   <none>        80/TCP   56m

The single pod is running and logs show no errors.

The rest of the cluster and services are working without issues.

Can anyone point me in a direction as to why my service seems to be unavailable?

Upvotes: 0

Views: 350

Answers (1)

Gijsriet
Gijsriet

Reputation: 163

In the logs of the aws-loadbalancer-controller I found:

"error":"ResourceInUse: Listener port '80' is in use by registered target

With the help of this article I learned you can share multiple ingresses with one ALB as long as you give them the same group name.

alb.ingress.kubernetes.io/group.name

So I moved all resources of the new app (Metabase) to a new namespace, added an extra ingress and to both ingresses added a group name like this:

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  annotations:
    alb.ingress.kubernetes.io/scheme: internet-facing
    alb.ingress.kubernetes.io/target-type: ip
    alb.ingress.kubernetes.io/group.name: dev

For me it did start up a new loadbalancer for the groupname, so I had to delete the old one.

Upvotes: 1

Related Questions