Merge pull request #2785 from BerriAI/litellm_high_traffic_redis_caching_fixes

[Feat] Proxy - high traffic redis caching - when using `url`

litellm/caching.py

@@ -109,6 +109,7 @@ class RedisCache(BaseCache):
         **kwargs,
     ):
         from ._redis import get_redis_client, get_redis_connection_pool
+        import redis
 
         redis_kwargs = {}
         if host is not None:
@@ -122,6 +123,14 @@ class RedisCache(BaseCache):
         self.redis_client = get_redis_client(**redis_kwargs)
         self.redis_kwargs = redis_kwargs
         self.async_redis_conn_pool = get_redis_connection_pool(**redis_kwargs)
+
+        if "url" in redis_kwargs and redis_kwargs["url"] is not None:
+            parsed_kwargs = redis.connection.parse_url(redis_kwargs["url"])
+            redis_kwargs.update(parsed_kwargs)
+            self.redis_kwargs.update(parsed_kwargs)
+            # pop url
+            self.redis_kwargs.pop("url")
+
         # redis namespaces
         self.namespace = namespace
         # for high traffic, we store the redis results in memory and then batch write to redis
@@ -338,6 +347,14 @@ class RedisCache(BaseCache):
                 traceback.print_exc()
                 raise e
 
+    def client_list(self):
+        client_list = self.redis_client.client_list()
+        return client_list
+
+    def info(self):
+        info = self.redis_client.info()
+        return info
+
     def flush_cache(self):
         self.redis_client.flushall()
 

litellm/proxy/proxy_server.py

@@ -8050,6 +8050,41 @@ async def cache_ping():
         )
 
 
+@router.get(
+    "/cache/redis/info",
+    tags=["caching"],
+    dependencies=[Depends(user_api_key_auth)],
+)
+async def cache_redis_info():
+    """
+    Endpoint for getting /redis/info
+    """
+    try:
+        if litellm.cache is None:
+            raise HTTPException(
+                status_code=503, detail="Cache not initialized. litellm.cache is None"
+            )
+        if litellm.cache.type == "redis":
+            client_list = litellm.cache.cache.client_list()
+            redis_info = litellm.cache.cache.info()
+            num_clients = len(client_list)
+            return {
+                "num_clients": num_clients,
+                "clients": client_list,
+                "info": redis_info,
+            }
+        else:
+            raise HTTPException(
+                status_code=500,
+                detail=f"Cache type {litellm.cache.type} does not support flushing",
+            )
+    except Exception as e:
+        raise HTTPException(
+            status_code=503,
+            detail=f"Service Unhealthy ({str(e)})",
+        )
+
+
 @router.post(
     "/cache/flushall",
     tags=["caching"],