vllm-sr 0.1.0b2.dev20260204090051__tar.gz → 0.1.0b2.dev20260204165623__tar.gz

{vllm_sr-0.1.0b2.dev20260204090051/vllm_sr.egg-info → vllm_sr-0.1.0b2.dev20260204165623}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: vllm-sr
-Version: 0.1.0b2.dev20260204090051
+Version: 0.1.0b2.dev20260204165623
 Summary: vLLM Semantic Router - Intelligent routing for Mixture-of-Models
 Author: vLLM-SR Team
 License: Apache-2.0

{vllm_sr-0.1.0b2.dev20260204090051 → vllm_sr-0.1.0b2.dev20260204165623}/cli/templates/envoy.template.yaml

@@ -36,8 +36,8 @@ static_resources:
                       exact: "{{ model.name }}"
                 route:
                   cluster: {{ model.cluster_name }}_cluster
-                  timeout: {{ listener.timeout | default('600s') }}
-                  idleTimeout: 600s
+                  timeout: {{ listener.timeout | default('1200s') }}
+                  idleTimeout: 1200s
                   # Rewrite Host header to match upstream server
                   host_rewrite_literal: "{{ model.endpoints[0].address }}"
                   {% if model.path_prefix %}
@@ -59,8 +59,8 @@ static_resources:
                       exact: "{{ model.name }}"
                 route:
                   cluster: anthropic_api_cluster
-                  timeout: {{ listener.timeout | default('600s') }}
-                  idleTimeout: 600s
+                  timeout: {{ listener.timeout | default('1200s') }}
+                  idleTimeout: 1200s
                   host_rewrite_literal: "api.anthropic.com"
               {% endfor %}
               # Default route (no x-selected-model header)
@@ -73,7 +73,7 @@ static_resources:
                   {% else %}
                   cluster: vllm_static_cluster
                   {% endif %}
-                  timeout: {{ listener.timeout | default('600s') }}
+                  timeout: {{ listener.timeout | default('1200s') }}
                   {% if models %}
                   # Rewrite Host header to match upstream server
                   host_rewrite_literal: "{{ models[0].endpoints[0].address }}"
@@ -94,13 +94,13 @@ static_resources:
               grpc_service:
                 envoy_grpc:
                   cluster_name: extproc_service
-                timeout: 600s
+                timeout: 1200s
               processing_mode:
                 request_header_mode: "SEND"
                 response_header_mode: "SEND"
                 request_body_mode: "BUFFERED"
                 response_body_mode: "BUFFERED"
-              message_timeout: {{ listener.timeout | default('600s') }}
+              message_timeout: {{ listener.timeout | default('1200s') }}
           - name: envoy.filters.http.router
             typed_config:
               "@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
@@ -115,7 +115,7 @@ static_resources:
   clusters:
   # ExtProc service (semantic router)
   - name: extproc_service
-    connect_timeout: 600s
+    connect_timeout: 1200s
     type: STATIC
     lb_policy: ROUND_ROBIN
     http2_protocol_options: {}
@@ -150,7 +150,7 @@ static_resources:
   {% for model in models %}
   # Cluster for model: {{ model.name }}
   - name: {{ model.cluster_name }}_cluster
-    connect_timeout: 600s
+    connect_timeout: 1200s
     type: {{ model.cluster_type }}
     {% if model.cluster_type == 'LOGICAL_DNS' %}
     dns_lookup_family: V4_ONLY
@@ -189,7 +189,7 @@ static_resources:
   - name: anthropic_api_cluster
     type: LOGICAL_DNS
     dns_lookup_family: V4_ONLY
-    connect_timeout: 600s
+    connect_timeout: 1200s
     lb_policy: ROUND_ROBIN
     load_assignment:
       cluster_name: anthropic_api_cluster

{vllm_sr-0.1.0b2.dev20260204090051 → vllm_sr-0.1.0b2.dev20260204165623}/cli/templates/router-defaults.yaml

@@ -173,7 +173,7 @@ looper:
   # Endpoint points to Envoy (same container), which handles load balancing and auth
   # Port should match listener port (default: 8888)
   endpoint: "http://localhost:8899/v1/chat/completions"
-  timeout_seconds: 120  # Timeout in seconds for each model call
+  timeout_seconds: 1200  # Timeout in seconds for each model call
   headers: {}  # Optional headers (e.g., {"Authorization": "Bearer xxx"})
 
 clear_route_cache: true

{vllm_sr-0.1.0b2.dev20260204090051 → vllm_sr-0.1.0b2.dev20260204165623}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "vllm-sr"
-version = "0.1.0.beta.2.dev20260204090051"
+version = "0.1.0.beta.2.dev20260204165623"
 description = "vLLM Semantic Router - Intelligent routing for Mixture-of-Models"
 authors = [{name = "vLLM-SR Team"}]
 readme = "README.md"

{vllm_sr-0.1.0b2.dev20260204090051 → vllm_sr-0.1.0b2.dev20260204165623/vllm_sr.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: vllm-sr
-Version: 0.1.0b2.dev20260204090051
+Version: 0.1.0b2.dev20260204165623
 Summary: vLLM Semantic Router - Intelligent routing for Mixture-of-Models
 Author: vLLM-SR Team
 License: Apache-2.0