fix: Add startupProbe configuration to order-service and payment-service

- Add startupProbe with 300s initialDelaySeconds and 60 failureThreshold to orderService
- Add startupProbe with 300s initialDelaySeconds and 60 failureThreshold to paymentService
- This matches gateway configuration and gives all Java services 5 minutes before first health check
- Total startup timeout is ~10 minutes to accommodate slow Spring application startup

k8s/helm/shop-recycle/values.yaml

@@ -189,6 +189,16 @@ orderService:
     timeoutSeconds: 3
     failureThreshold: 3
   
+  # 启动探针 - 给 Java 应用充分的启动时间（5分钟才开始检查）
+  startupProbe:
+    httpGet:
+      path: /health
+      port: 8081
+    initialDelaySeconds: 300
+    periodSeconds: 10
+    timeoutSeconds: 5
+    failureThreshold: 60
+
   service:
     type: ClusterIP
     port: 8081
@@ -242,6 +252,16 @@ paymentService:
     periodSeconds: 5
     timeoutSeconds: 3
     failureThreshold: 3
+  # 启动探针 - 给 Java 应用充分的启动时间（5分钟才开始检查）
+  startupProbe:
+    httpGet:
+      path: /health
+      port: 8082
+    initialDelaySeconds: 300
+    periodSeconds: 10
+    timeoutSeconds: 5
+    failureThreshold: 60
+
   
   service:
     type: ClusterIP