neuralbank-stack

neuralbank-stack Helm Chart

A Helm chart to deploy the NeuralBank stack: PostgreSQL (configured via globals), backend, frontend and an NGINX proxy/gateway.

Contents

• backend: API service (configurable image, port, replicas)
• frontend: static/web app (configurable image, port, replicas)
• proxy: nginx gateway (LoadBalancer/NodePort)
• global.postgresql: external or chart-provided PostgreSQL credentials

Requirements

• Kubernetes cluster (v1.20+ recommended)
• Helm 3+
• A PostgreSQL instance or a PostgreSQL chart that sets:
  • global.postgresql.auth.postgresPassword
  • global.postgresql.auth.username
  • global.postgresql.auth.password
  • global.postgresql.auth.database

If you prefer an external DB, set those globals to point to your DB credentials.

Values (high level)

Key values from values.yaml:

• global.postgresql.auth
  • postgresPassword, username, password, database
• backend
  • image.repository
  • image.tag
  • image.pullPolicy
  • service.port (default 8080)
  • replicaCount
• frontend
  • image.repository
  • image.tag
  • image.pullPolicy
  • service.port (default 80)
  • replicaCount
• proxy
  • image.repository
  • image.tag

  • service.type (LoadBalancer

    NodePort)

  • service.port (default 80)

See values.yaml for the full defaults and comments.

Installation

1. From the chart directory:

   helm install neuralbank . \
     --namespace neuralbank \
     --create-namespace \
     -f values.yaml
   

2. To use a custom values file:

   helm install neuralbank . -f my-values.yaml --namespace neuralbank --create-namespace
   

If using an existing PostgreSQL, ensure you set the globals in your values file:

global:
  postgresql:
    auth:
      postgresPassword: "supersecretpassword"
      username: "neuralbank"
      password: "neuralbank123!"
      database: "neuralbank"

Upgrading

helm upgrade neuralbank . -f values.yaml --namespace neuralbank

Uninstall

helm uninstall neuralbank --namespace neuralbank
kubectl delete namespace neuralbank

Accessing services

• Backend container port: 8080 (service listens on values.backend.service.port)
• Frontend container port: 80
• Proxy: default LoadBalancer on port 80 (or NodePort when configured)

Run:

kubectl get svc -n neuralbank
kubectl get pods -n neuralbank

Notes

• Images default to quay.io/… with tag “latest”. Set explicit tags in values.yaml for reproducible deployments.
• No persistent volumes are defined by default in values.yaml. Add persistence settings if required.
• For local testing without a LoadBalancer, switch proxy.service.type to NodePort.

Maintainers

• Repository owner: Maximiliano Pizarro