cleanup all worker stats when starting up servers

server/Server.py

@@ -25,7 +25,7 @@ class Server:
     'version': '1.0.0',
   }
 
-  def __init__(self, data_path:str, snake_type:str, storage_type:str, debug:bool=False, check_tls_security:bool=False, game_state_backend:str='memory', game_state_redis_url:str='redis://localhost:6379/0', game_state_ttl_sec:int=900, game_state_local_cache:bool=True, metrics_backend:str='memory', metrics_redis_url:str='redis://localhost:6379/0', metrics_ttl_sec:int=None):
+  def __init__(self, data_path:str, snake_type:str, storage_type:str, debug:bool=False, check_tls_security:bool=False, game_state_backend:str='memory', game_state_redis_url:str='redis://localhost:6379/0', game_state_ttl_sec:int=900, game_state_local_cache:bool=True, metrics_backend:str='memory', metrics_redis_url:str='redis://localhost:6379/0', metrics_ttl_sec:int|None=None):
     self.debug = debug
     self.snake_type = snake_type
     self.storage_type = storage_type
@@ -44,9 +44,11 @@ class Server:
     )
     metrics_backend_normalized = (metrics_backend or 'memory').strip().lower()
     self.stale_game_timeout_sec = self._get_stale_game_timeout_sec()
+
     self.running_games:dict[str, GameBoard] = {}
     self.game_move_counts:dict[str, int] = {}
     self.game_last_seen_unix:dict[str, int] = {}
+
     self.metrics_collector = ServerMetricsCollector(
       metrics_manager=MetricsManager(
         backend=metrics_backend_normalized,
@@ -61,6 +63,10 @@ class Server:
       game_last_seen_unix=self.game_last_seen_unix,
       game_move_counts=self.game_move_counts,
     )
+    self.clear_worker_metrics_on_startup = self._env_bool('METRICS_CLEAR_WORKERS_ON_STARTUP', True)
+    self.worker_metrics_startup_lock_ttl_sec = self._env_int('METRICS_STARTUP_CLEANUP_LOCK_TTL_SEC', 300)
+    self._startup_worker_metrics_cleared = False
+
     self.logger = build_logger('Battlesnake', debug_env_var='DEBUG_SERVER')
     self.snake_version = self._get_snake_version()
 
@@ -136,6 +142,16 @@ class Server:
       response.headers.set('server', 'battlesnake/gitea/snake-python')
       return response
 
+    @self.app.before_serving
+    async def clear_startup_worker_metrics_once():
+      if self._startup_worker_metrics_cleared:
+        return
+      self._startup_worker_metrics_cleared = True
+      if self.clear_worker_metrics_on_startup:
+        should_clear = await self.metrics_collector.should_clear_worker_metrics_on_startup(self.worker_metrics_startup_lock_ttl_sec)
+        if should_clear:
+          await self.metrics_collector.clear_worker_metrics()
+
     @self.app.after_serving
     async def shutdown_state_storage():
       await self.game_state_store.close()
@@ -210,6 +226,21 @@ class Server:
     except ValueError:
       return 180
 
+  def _env_bool(self, name:str, default:bool=False) -> bool:
+    value = os.getenv(name)
+    if value is None:
+      return default
+    return value.strip().lower() in {'1', 'true', 'yes', 'on'}
+
+  def _env_int(self, name: str, default: int) -> int:
+    value = os.getenv(name)
+    if value is None:
+      return default
+    try:
+      return int(value)
+    except ValueError:
+      return default
+
   async def _create_game_board(self, game_state:dict) -> GameBoard:
     game_id = game_state['game']['id']
     new_game_board = GameBoard(