Try to fix connectivity issues

deploy/prod-pod.yml

@@ -73,7 +73,7 @@ spec:
         drop: ["ALL"]
     env:
     - name: DATABASE_URL
-      value: "postgres://${POSTGRES_USERNAME}:${POSTGRES_PASSWORD}@localhost:${POSTGRES_PORT}/${POSTGRES_DATABASE_NAME}?sslmode=disable"
+      value: "postgres://${POSTGRES_USERNAME}:${POSTGRES_PASSWORD}@127.0.0.1:${POSTGRES_PORT}/${POSTGRES_DATABASE_NAME}?sslmode=disable"
     - name: PORT
       value: "${PROD_BACKEND_PORT}"
     ports:
@@ -162,7 +162,6 @@ spec:
       runAsGroup: 1000
     ports:
       - containerPort: 8080           # inside pod
-      - containerPort: 8090           # health inside pod (not exposed)
     volumeMounts:
       - { name: nginx-run,   mountPath: /var/run,              readOnly: false }
       - { name: nginx-cache, mountPath: /var/cache/nginx,      readOnly: false }
@@ -171,7 +170,7 @@ spec:
         mountPath: /tmp
     # Health check
     livenessProbe:
-      httpGet: { path: /healthz, port: 8090, scheme: HTTP }
+      httpGet: { path: /healthz, port: 8080, scheme: HTTP }
       initialDelaySeconds: 10
       periodSeconds: 30
     # Resource limits

nginx/nginx.conf

@@ -15,16 +15,17 @@ http {
   uwsgi_temp_path /tmp/uwsgi_temp;
   scgi_temp_path /tmp/scgi_temp;
 
-  # health
-  server {
-    listen 8090;
-    location = /healthz { return 200 "ok\n"; add_header Content-Type text/plain; }
-  }
-
   # public HTTP entrypoint (host will terminate TLS and proxy here)
     server {
     listen 8080;
 
+    # exact-match: wins over everything else in this server
+    location = /healthz {
+        access_log off;
+        add_header Content-Type text/plain;
+        return 200 "ok\n";
+    }
+
     # frontend default
     location / {
         proxy_http_version 1.1;