Speed up rolling deploys: allow surge, reduce startup delay

- maxSurge 0→1: new pods start before old ones terminate
- minReadySeconds 10→5
- startupProbe: initialDelay 10→5s, period 5→3s

k8s/deployment.yaml

@@ -5,11 +5,11 @@ metadata:
   namespace: prop
 spec:
   replicas: 3
-  minReadySeconds: 10
+  minReadySeconds: 5
   strategy:
     type: RollingUpdate
     rollingUpdate:
-      maxSurge: 0
+      maxSurge: 1
       maxUnavailable: 1
   selector:
     matchLabels:
@@ -67,9 +67,9 @@ spec:
             httpGet:
               path: /health
               port: 5000
-            initialDelaySeconds: 10
+            initialDelaySeconds: 5
-            periodSeconds: 5
+            periodSeconds: 3
-            failureThreshold: 12
+            failureThreshold: 10
             timeoutSeconds: 3
           readinessProbe:
             httpGet: