fix: Adjust frontEnd thrasholds (#2288)

Author: atchernych

components/backends/sglang/deploy/agg.yaml

@@ -12,10 +12,10 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command:

components/backends/sglang/deploy/agg_router.yaml

@@ -12,10 +12,10 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command:

components/backends/sglang/deploy/disagg.yaml

@@ -12,10 +12,10 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command:

components/backends/vllm/deploy/agg.yaml

@@ -12,20 +12,20 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command:
             - /bin/sh
             - -c
             - 'curl -s http://localhost:8000/health | jq -e ".status == \"healthy\""'
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 10
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       dynamoNamespace: vllm-agg
       componentType: main
       replicas: 1
@@ -51,15 +51,17 @@ spec:
         httpGet:
           path: /live
           port: 9090
-        periodSeconds: 5
-        timeoutSeconds: 30
-        failureThreshold: 1
+        initialDelaySeconds: 30
+        periodSeconds: 10
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         httpGet:
           path: /health
           port: 9090
+        initialDelaySeconds: 30
         periodSeconds: 10
-        timeoutSeconds: 30
+        timeoutSeconds: 5
         failureThreshold: 60
       dynamoNamespace: vllm-agg
       componentType: worker
@@ -86,7 +88,9 @@ spec:
             httpGet:
               path: /health
               port: 9090
+            initialDelaySeconds: 10
             periodSeconds: 10
+            timeoutSeconds: 5
             failureThreshold: 60
           image: nvcr.io/nvidian/nim-llm-dev/vllm-runtime:dep-233.17
           workingDir: /workspace/components/backends/vllm

components/backends/vllm/deploy/agg_router.yaml

@@ -12,10 +12,10 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command:

components/backends/vllm/deploy/disagg.yaml

@@ -15,10 +15,10 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command:

components/backends/vllm/deploy/disagg_planner.yaml

@@ -22,10 +22,10 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command:

components/backends/vllm/deploy/disagg_router.yaml

@@ -15,10 +15,10 @@ spec:
         httpGet:
           path: /health
           port: 8000
-        initialDelaySeconds: 60
-        periodSeconds: 60
-        timeoutSeconds: 30
-        failureThreshold: 10
+        initialDelaySeconds: 20
+        periodSeconds: 5
+        timeoutSeconds: 5
+        failureThreshold: 3
       readinessProbe:
         exec:
           command: