fix: resolve critical deployment issues across Helm charts, Dockerfile, and K8s fallback

- Add helm/kubectl binaries and chart directory to backend Dockerfile
- Extend Helm templates for MongoDB/MariaDB database support (env vars, probes, ports)
- Add Redis and RabbitMQ Helm templates (deployment, service, secret, PVC)
- Add generic app-storage PVC and Fluent Bit sidecar with ES authentication
- Fix imagePullSecrets in K8s API fallback, prevent secret regeneration on redeploy
- Clean up Redis/RabbitMQ/FluentBit resources on app deletion without removing shared secrets
- Fix HelmService chartPath resolution for production Docker builds

Co-authored-by: Cursor <cursoragent@cursor.com>

backend/helm/cloudhost-app/values.yaml

@@ -7,10 +7,11 @@
 app:
   name: my-app
   namespace: user-default
-  runtime: nodejs          # nodejs | laravel | wordpress
+  runtime: nodejs          # nodejs | laravel | wordpress | go | php | python | django | dotnet
   image: ""                # e.g. registry.example.com/my-app:v1
   port: 3000
   replicas: 1
+  storageSize: "2Gi"       # PVC size for app files (uploads, media, data)
 
 # ── Resources ────────────────────────────────────────────
 resources:
@@ -51,6 +52,33 @@ wordpress:
   enabled: false
   wpContentStorageSize: "2Gi"
 
+# ── Redis ──────────────────────────────────────────────
+redis:
+  enabled: false
+  version: "7.2"
+  storageSize: "1Gi"
+  resources:
+    cpuRequest: "50m"
+    cpuLimit: "200m"
+    memoryRequest: "64Mi"
+    memoryLimit: "256Mi"
+
+# ── RabbitMQ ──────────────────────────────────────────
+rabbitmq:
+  enabled: false
+  version: "3.13"
+  storageSize: "2Gi"
+  resources:
+    cpuRequest: "100m"
+    cpuLimit: "500m"
+    memoryRequest: "256Mi"
+    memoryLimit: "512Mi"
+
+# ── Elasticsearch (Fluent Bit sidecar for logging) ────
+elasticsearch:
+  enabled: false
+  logPaths: []
+
 # ── Change metadata ─────────────────────────────────────
 changeCause: ""