Ingress backend shows unhealthy in GCP

Question

I am trying to expose my application on classic ALB on GCP through ingress. I've created a Nodeport service below:

apiVersion: v1
kind: Service
metadata:
  name: ofamily-service-np
spec:
  selector:
    app: o-famil
  type: NodePort
  ports:
    - protocol: TCP
      port: 8000
      targetPort: 8000

I have also created managedCertificate object:

apiVersion: networking.gke.io/v1
kind: ManagedCertificate
metadata:
  name: duckdns
spec:
  domains:
    - ofamily.duckdns.org

Here's my ingress.yaml

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: o-family-ingress
  annotations:
    kubernetes.io/ingress.class: "gce"
    networking.gke.io/managed-certificates: "duckdns"
    kubernetes.io/ingress.global-static-ip-name: "ingress"
spec:
  rules:
  - host: ofamily.duckdns.org
    http:
      paths:
      - pathType: Prefix
        path: "/"
        backend:
          service:
            name: ofamily-service-np
            port:
              number: 8000

It creates a LoadBalancer

in the LoadBalancer it shows Network endpoint group details which shows unhealthy, I am unable to troubleshoot.

I am trying to expose my application on classic ALB on GCP through ingress. I've created a Nodeport service and ingress.yaml but the backend shows unhealthy. I am using GKE Autopilot cluster.

The same application works on loadbalancer service type but when I use Nodeport and ingress, it shows unhealthy.

Do I need to change something in the application or add readiness/liveliness probe in deployment or ingress?

ingress-error1ingress-error2