[model-gateway] Add PostgreSQL support to binding (#13766)

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Simo Lin <linsimo.mark@gmail.com>
Co-authored-by: Chang Su <chang.s.su@oracle.com>

docs/advanced_features/router.md

@@ -275,6 +275,7 @@ PD deployments can specify `--prefill-selector` and `--decode-selector` plus the
 | `memory` (default) | In-memory storage for quick prototyping. | `--history-backend memory` |
 | `none` | No persistence; APIs operate but store nothing. | `--history-backend none` |
 | `oracle` | Oracle Autonomous Database-backed storage (pooled connections). | `--history-backend oracle` |
+| `postgres` | PostgreSQL Database-backed storage (pooled connections). | `--history-backend postgres` |
 
 Oracle configuration (choose DSN *or* TNS alias):
 Install the Oracle Instant Client and set `LD_LIBRARY_PATH` accordingly.

sgl-router/README.md

@@ -454,6 +454,7 @@ Public health endpoints (`/liveness`, `/readiness`, `/health`, `/health_generate
 - `--history-backend memory` (default) stores responses and conversations in-process.
 - `--history-backend none` disables persistence while keeping APIs.
 - `--history-backend oracle` uses Oracle Autonomous Database; provide credentials via flags or environment variables.
+- `--history-backend postgres` uses PostgreSQL Database.
 - Conversation item storage mirrors the history backend (Oracle or memory). The same storage powers OpenAI `/responses` and conversation APIs.
 
 ### History Backend (OpenAI Router Mode)
@@ -465,6 +466,7 @@ Store conversation and response data for tracking, debugging, or analytics.
 - **Memory** (default): In-memory storage, fast but ephemeral.
 - **None**: No storage, minimal overhead.
 - **Oracle**: Persistent storage backed by Oracle Autonomous Database.
+- **Postgres**: Persistent storage backed by PostgreSQL Database.
 
 ```bash
 # Memory backend (default)
@@ -484,6 +486,12 @@ python3 -m sglang_router.launch_router \
   --backend openai \
   --worker-urls https://api.openai.com \
   --history-backend oracle
+
+# PostgreSQL backend
+python3 -m sglang_router.launch_router \
+  --backend openai \
+  --worker-urls https://api.openai.com \
+  --history-backend postgres
 ```
 
 #### Oracle configuration

sgl-router/bindings/python/sglang_router/router.py

@@ -210,6 +210,8 @@ class Router:
             "oracle_pool_min",
             "oracle_pool_max",
             "oracle_pool_timeout_secs",
+            "postgres_db_url",
+            "postgres_pool_max",
         ]
         for field in fields_to_remove:
             args_dict.pop(field, None)

sgl-router/bindings/python/sglang_router/router_args.py

@@ -109,6 +109,8 @@ class RouterArgs:
     oracle_pool_min: int = 1
     oracle_pool_max: int = 16
     oracle_pool_timeout_secs: int = 30
+    postgres_db_url: Optional[str] = None
+    postgres_pool_max: int = 16
     # mTLS configuration for worker communication
     client_cert_path: Optional[str] = None
     client_key_path: Optional[str] = None
@@ -598,6 +600,19 @@ class RouterArgs:
             ),
             help="Oracle connection pool timeout in seconds (default: 30, env: ATP_POOL_TIMEOUT_SECS)",
         )
+        # Postgres configuration
+        parser.add_argument(
+            f"--{prefix}postgres-db-url",
+            type=str,
+            default=os.getenv("POSTGRES_DB_URL"),
+            help="PostgreSQL database connection URL (env: POSTGRES_DB_URL)",
+        )
+        parser.add_argument(
+            f"--{prefix}postgres-pool-max",
+            type=int,
+            default=int(os.getenv("POSTGRES_POOL_MAX", RouterArgs.postgres_pool_max)),
+            help="Maximum PostgreSQL connection pool size (default: 16, env: POSTGRES_POOL_MAX)",
+        )
         # mTLS configuration
         parser.add_argument(
             f"--{prefix}client-cert-path",