feat: parameterize ports/pod, fix valkey healthcheck, address all review comments

.env.dev

@@ -0,0 +1,27 @@
+# Dev environment overlay — sourced AFTER .env to override prod values
+# Secrets are inherited from .env; only env-specific config is set here.
+
+# Pod identity
+POD_NAME="dev-router-pod"
+
+# Public URLs
+BASEURL="dev.vendeuvre.lan"
+BASE_URL="dev.vendeuvre.lan"
+PUBLIC_BASE_URL="https://dev.vendeuvre.lan/llm-routing"
+
+# Dev port assignments (+10 offset from production)
+ROUTER_PORT="5010"
+LITELLM_PORT="4010"
+LANGFUSE_WEB_PORT="3011"
+LANGFUSE_WORKER_PORT="3040"
+POSTGRES_PORT="5442"
+VALKEY_CACHE_PORT="6389"
+VALKEY_LF_PORT="6390"
+CLICKHOUSE_HTTP_PORT="8133"
+CLICKHOUSE_TCP_PORT="9010"
+CLICKHOUSE_INTERSERVER_PORT="9019"
+MINIO_S3_PORT="9012"
+MINIO_CONSOLE_PORT="9011"
+
+# Dev uses locally-built router image (for testing code changes)
+ROUTER_IMAGE="localhost/llm-routing-dev:latest"

.gitignore

@@ -4,13 +4,9 @@
 # Exclude Podman runtime file generated dynamically at startup
 pod-run.yaml
 
-# Exclude persistent database and cache storages
-postgres-data/
-langfuse-data/
-valkey-data/
-clickhouse-data/
-minio-data/
-redis-lf-data/
+# Exclude persistent database and cache storages (all under data/)
+data/
+
 
 # Exclude auto-generated backups
 backups/
@@ -36,7 +32,8 @@ router/free_models_roster.json
 workdirs/
 pr_description.txt
 
-# Dataset work in progress
-data/
+# Python virtual environment
+.venv/
+
 .cache/
 test_output*.log

README.md

@@ -76,12 +76,12 @@ All core containers are configured with **Kubernetes-style liveness and readines
 
 | Container | Liveness Probe | Readiness Probe |
 |:---|---:|---:|
-| **valkey-cache** | `tcpSocket` on port 6379 every 10s | Same, every 5s |
+| **valkey-cache** | `valkey-cli -p <port> ping` every 10s | Same, every 5s |
 | **litellm-gateway** | Python `urllib` GET `/ping` (port 4000) every 15s | Python `urllib` GET `/health/readiness` (port 4000) every 10s |
 | **llm-triage-router** | Python `urllib` GET `/metrics` (port 5000) every 15s | Same, every 10s |
-| **postgres-db** | `pg_isready -U postgres` every 10s | Same, every 5s |
+| **postgres-db** | `pg_isready -U postgres -p <port>` every 10s | Same, every 5s |
 | **clickhouse-db** | `clickhouse-client --user clickhouse --password <generated> --query "SELECT 1"` every 15s | `clickhouse-client --user clickhouse --password <generated> --query "SELECT 1"` every 10s |
-| **valkey-lf** | `tcpSocket` on port 6380 every 10s | Same, every 5s |
+| **valkey-lf** | `valkey-cli -p <port> -a <auth> ping` every 10s | Same, every 5s |
 | **langfuse-web** | `wget` GET `/api/health` (port 3001) every 15s | Same, every 10s |
 | **langfuse-worker** | `pgrep node` every 15s | — |
 | **minio-s3** | `httpGet` `/minio/health/live` (port 9002) every 15s | `httpGet` `/minio/health/ready` (port 9002) every 10s |

litellm/config.yaml

@@ -12,7 +12,7 @@ litellm_settings:
   cache_params:
     host: 127.0.0.1
     mode: default
-    port: 6379
+    port: VALKEY_CACHE_PORT_PLACEHOLDER
     ttl: 3600
     type: redis
   caching_backend: redis
@@ -77,7 +77,7 @@ model_list:
     is_public_model_group: true
 - litellm_params:
     model: openai/llm-routing-ollama
-    api_base: http://127.0.0.1:5000/v1
+    api_base: http://127.0.0.1:ROUTER_PORT_PLACEHOLDER/v1
     api_key: os.environ/LITELLM_MASTER_KEY
     request_timeout: 120
   model_name: llm-routing-ollama
@@ -237,7 +237,7 @@ model_list:
 
 redis_settings:
   redis_host: 127.0.0.1
-  redis_port: 6379
+  redis_port: VALKEY_CACHE_PORT_PLACEHOLDER
 router_settings:
   allowed_fails: 0
   cooldown_time: 300

litellm/entrypoint.py

@@ -16,8 +16,9 @@
             line = line.strip()
             if line and not line.startswith("#") and "=" in line:
                 key, _, val = line.partition("=")
+                key = key.strip()
                 val = val.strip().strip('"').strip("'")
-                os.environ[key] = val
+                os.environ.setdefault(key, val)
 
 # Load Gemini OAuth token from credentials JSON
 creds_path = "/config/gemini_auth/oauth_creds.json"
@@ -46,9 +47,15 @@ def check_tcp_port(ip: str, port: int) -> bool:
         return False
 
 max_wait = 60
-print(f"🔌 Waiting for PostgreSQL on :5432 (max {max_wait}s)...")
+postgres_port_str = os.environ.get("POSTGRES_PORT") or "5432"
+try:
+    postgres_port = int(postgres_port_str)
+except ValueError:
+    print(f"⚠️ Invalid POSTGRES_PORT '{postgres_port_str}', defaulting to 5432")
+    postgres_port = 5432
+print(f"🔌 Waiting for PostgreSQL on :{postgres_port} (max {max_wait}s)...")
 for i in range(max_wait):
-    if check_tcp_port("127.0.0.1", 5432):
+    if check_tcp_port("127.0.0.1", postgres_port):
         print(f"✅ PostgreSQL ready after {i+1}s")
         break
     time.sleep(1)
@@ -127,5 +134,6 @@ def _serialize_dt(dt):
 # Start LiteLLM Proxy
 import litellm
 from litellm.proxy.proxy_cli import run_server
-sys.argv = ["litellm", "--config", "/app/config.yaml", "--port", "4000"]
+litellm_port = os.environ.get("LITELLM_PORT") or os.environ.get("PORT") or "4000"
+sys.argv = ["litellm", "--config", "/app/config.yaml", "--port", litellm_port]
 run_server()