use ingress/egress

2026-06-17 15:25:17 +02:00 · 2025-02-14 16:43:51 -08:00 · 2025-02-14 16:43:51 -08:00 · 136daa2d3c
commit 136daa2d3c
parent 4a957f2b86
18 changed files with 31 additions and 27 deletions
--- a/arch/arch_config_schema.yaml
+++ b/arch/arch_config_schema.yaml
@ -5,8 +5,9 @@ properties:
    type: string
  listeners:
    type: object
+    additionalProperties: false
    properties:
-      prompt_gateway:
+      ingress_traffic:
        type: object
        properties:
          address:
@ -20,7 +21,7 @@ properties:
          timeout:
            type: string
        additionalProperties: false
-      llm_gateway:
+      egress_traffic:
        type: object
        properties:
          address:
@ -31,7 +32,6 @@ properties:
            type: string
            enum:
              - openai
-              - huggingface
          timeout:
            type: string
        additionalProperties: false
@ -247,5 +247,4 @@ properties:
 additionalProperties: false
 required:
  - version
-  - listeners
  - llm_providers
--- a/arch/tools/cli/config_generator.py
+++ b/arch/tools/cli/config_generator.py
@ -104,7 +104,9 @@ def validate_and_render_schema():
    arch_config_string = yaml.dump(config_yaml)
    arch_llm_config_string = yaml.dump(config_yaml)

-    prompt_gateway_listener = config_yaml.get("listeners", {}).get("prompt_gateway", {})
+    prompt_gateway_listener = config_yaml.get("listeners", {}).get(
+        "ingress_traffic", {}
+    )
    if prompt_gateway_listener.get("port") == None:
        prompt_gateway_listener["port"] = 10000  # default port for prompt gateway
    if prompt_gateway_listener.get("address") == None:
@ -112,7 +114,7 @@ def validate_and_render_schema():
    if prompt_gateway_listener.get("timeout") == None:
        prompt_gateway_listener["timeout"] = "10s"

-    llm_gateway_listener = config_yaml.get("listeners", {}).get("llm_gateway", {})
+    llm_gateway_listener = config_yaml.get("listeners", {}).get("egress_traffic", {})
    if llm_gateway_listener.get("port") == None:
        llm_gateway_listener["port"] = 12000  # default port for llm gateway
    if llm_gateway_listener.get("address") == None:
--- a/arch/tools/cli/core.py
+++ b/arch/tools/cli/core.py
@ -143,12 +143,12 @@ def start_arch(arch_config_file, env, log_timeout=120, foreground=False):

        prompt_gateway_port = (
            arch_config_dict.get("listeners", {})
-            .get("prompt_gateway", {})
+            .get("ingress_traffic", {})
            .get("port", 10000)
        )
        llm_gateway_port = (
            arch_config_dict.get("listeners", {})
-            .get("llm_gateway", {})
+            .get("egress_traffic", {})
            .get("port", 12000)
        )

--- a/demos/samples_java/weather_forcecast_service/arch_config.yaml
+++ b/demos/samples_java/weather_forcecast_service/arch_config.yaml
@ -1,6 +1,6 @@
 version: v0.1
 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/demos/samples_python/currency_exchange/arch_config.yaml
+++ b/demos/samples_python/currency_exchange/arch_config.yaml
@ -1,7 +1,7 @@
 version: v0.1

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/demos/samples_python/human_resources_agent/arch_config.yaml
+++ b/demos/samples_python/human_resources_agent/arch_config.yaml
@ -1,6 +1,6 @@
 version: v0.1
 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/demos/samples_python/multi_turn_rag_agent/arch_config.yaml
+++ b/demos/samples_python/multi_turn_rag_agent/arch_config.yaml
@ -1,7 +1,7 @@
 version: v0.1

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/demos/samples_python/network_switch_operator_agent/arch_config.yaml
+++ b/demos/samples_python/network_switch_operator_agent/arch_config.yaml
@ -1,6 +1,6 @@
 version: v0.1
 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/demos/samples_python/stock_quote/arch_config.yaml
+++ b/demos/samples_python/stock_quote/arch_config.yaml
@ -1,11 +1,12 @@
 version: v0.1

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
    timeout: 30s
+
 llm_providers:
  - name: gpt-4o
    access_key: $OPENAI_API_KEY
--- a/demos/samples_python/weather_forecast/arch_config.yaml
+++ b/demos/samples_python/weather_forecast/arch_config.yaml
@ -1,7 +1,7 @@
 version: "0.1-beta"

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/demos/use_cases/llm_routing/arch_config.yaml
+++ b/demos/use_cases/llm_routing/arch_config.yaml
@ -1,9 +1,9 @@
 version: "0.1-beta"

 listeners:
-  prompt_gateway:
+  egress_traffic:
    address: 0.0.0.0
-    port: 10000
+    port: 12000
    message_format: openai
    timeout: 30s

--- a/demos/use_cases/ollama/arch_config.yaml
+++ b/demos/use_cases/ollama/arch_config.yaml
@ -1,7 +1,7 @@
 version: v0.1

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/demos/use_cases/spotify_bearer_auth/arch_config.yaml
+++ b/demos/use_cases/spotify_bearer_auth/arch_config.yaml
@ -1,6 +1,6 @@
 version: v0.1
 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/docs/source/concepts/includes/arch_config.yaml
+++ b/docs/source/concepts/includes/arch_config.yaml
@ -1,7 +1,7 @@
 version: v0.1

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/docs/source/get_started/quickstart.rst
+++ b/docs/source/get_started/quickstart.rst
@ -43,7 +43,7 @@ Create ``arch_config.yaml`` file with the following content:
   version: v0.1

  listeners:
-    prompt_gateway:
+    ingress_traffic:
      address: 0.0.0.0
      port: 10000
      message_format: openai
@ -146,9 +146,9 @@ Create ``arch_config.yaml`` file with the following content:
   version: v0.1

  listeners:
-    prompt_gateway:
+    egress_traffic:
      address: 0.0.0.0
-      port: 10000
+      port: 12000
      message_format: openai
      timeout: 30s

--- a/docs/source/guides/includes/arch_config.yaml
+++ b/docs/source/guides/includes/arch_config.yaml
@ -1,7 +1,7 @@
 version: v0.1

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
--- a/docs/source/resources/includes/arch_config_full_reference.yaml
+++ b/docs/source/resources/includes/arch_config_full_reference.yaml
@ -1,14 +1,16 @@
 version: v0.1

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai
    timeout: 5s
-  llm_gateway:
+  egress_traffic:
    address: 0.0.0.0
    port: 12000
+    message_format: openai
+    timeout: 5s

 # Arch creates a round-robin load balancing between different endpoints, managed via the cluster subsystem.
 endpoints:
--- a/tests/archgw/arch_config.yaml
+++ b/tests/archgw/arch_config.yaml
@ -1,7 +1,7 @@
 version: "0.1-beta"

 listeners:
-  prompt_gateway:
+  ingress_traffic:
    address: 0.0.0.0
    port: 10000
    message_format: openai