[feat] add metrics for yiyan adapter (#3219) (#3614)

* [feat] add metrics for yiyan adapter

* [fix] fix metrics num_requests_waiting and num_requests_running

* [fix] fix metrics gpu_cache_usage_perc

* [refactor] change where requests_number increases

* [chore] rename xxx_block_num as xxx_gpu_block_num, and update their values accordingly

* [chore] delete useless code

fastdeploy/engine/common_engine.py

@@ -552,6 +552,8 @@ class EngineSevice:
                     get_request_pool.submit(_fetch_request)
                 # 2. Schedule requests
                 tasks = self.resource_manager.schedule()
+                main_process_metrics.num_requests_waiting.dec(len(tasks))
+                main_process_metrics.num_requests_running.inc(len(tasks))
                 # 3. Send to engine
                 if tasks:
                     self.resource_manager.get_real_bsz()
@@ -597,6 +599,7 @@ class EngineSevice:
                     try:
                         request = Request.from_dict(data)
                         start_span("ENQUEUE_ZMQ", data, trace.SpanKind.PRODUCER)
+                        main_process_metrics.requests_number.inc()
                         llm_logger.debug(f"Receive request: {request}")
                     except Exception as e:
                         llm_logger.error(f"Receive request error: {e}, {traceback.format_exc()!s}")