deploy-dev

name	deploy-dev
description	Deploy services to the dev EKS namespace via ArgoCD (or kubectl for bootstrap)
disable-model-invocation	true
argument-hint	[service|all]

/deploy-dev [service|all]

Deploy services to the petclinic-dev namespace.

Arguments

• service — Specific service to deploy (e.g., config-server, api-gateway) or all
• Default: all

Steps

1. Verify kubectl context is configured for the dev cluster:

   kubectl config current-context
   

   If not configured, run:

   aws eks update-kubeconfig --name petclinic-dev --region eu-central-1
   

2. Check if ArgoCD is installed:

   kubectl get namespace argocd 2>/dev/null
   

3. If ArgoCD is installed (standard path):

   • all: Sync all dev ArgoCD Applications:

     argocd app sync -l environment=dev
     

   • specific service: Sync that service's ArgoCD Application:

     argocd app sync {service}-dev
     

   • Wait for sync to complete:

     argocd app wait {service}-dev --timeout 120
     

4. If ArgoCD is NOT installed (bootstrap path):

   • Apply namespace first (idempotent):

     kubectl apply -f k8s/base/namespaces.yaml
     

   • all: Install via Helm directly:

     for service in config-server discovery-server api-gateway customers-service visits-service vets-service genai-service admin-server; do
       helm upgrade --install $service helm/petclinic-service/ \
         -f helm-values/$service.yaml \
         -f helm-values/dev.yaml \
         -n petclinic-dev --create-namespace
     done
     

   • specific service: Install that service via Helm:

     helm upgrade --install {service} helm/petclinic-service/ \
       -f helm-values/{service}.yaml \
       -f helm-values/dev.yaml \
       -n petclinic-dev
     

5. For service startup order (when deploying all):

   • Deploy config-server first, wait for ready
   • Deploy discovery-server second, wait for ready
   • Deploy all remaining services

6. Monitor rollout status:

   kubectl rollout status deployment/{service} -n petclinic-dev --timeout=120s
   

7. Show final pod status:

   kubectl get pods -n petclinic-dev -o wide
   

8. If any pod fails to start, show logs:

   kubectl logs deployment/{failing-service} -n petclinic-dev --tail=50
   

Important

• This deploys to DEV only — use /deploy-prod for production
• Prefer ArgoCD sync when available (standard path after E-17)
• Monitor the rollout — don't assume success without checking pod status
• If a deployment fails, show the logs and describe the issue