Refactor runner and API components to remove IP address handling. Update client and server logic to streamline runner registration and task distribution. Introduce write mutexes for connection management to enhance concurrency control. Clean up whitespace and improve code readability across multiple files.

internal/api/server.go

@@ -38,6 +38,9 @@ type Server struct {
 	wsUpgrader      websocket.Upgrader
 	runnerConns     map[int64]*websocket.Conn
 	runnerConnsMu   sync.RWMutex
+	// Mutexes for each runner connection to serialize writes
+	runnerConnsWriteMu   map[int64]*sync.Mutex
+	runnerConnsWriteMuMu sync.RWMutex
 	frontendConns   map[string]*websocket.Conn // key: "jobId:taskId"
 	frontendConnsMu sync.RWMutex
 	// Mutexes for each frontend connection to serialize writes
@@ -55,6 +58,8 @@ type Server struct {
 	// Throttling for progress updates (per job)
 	progressUpdateTimes   map[int64]time.Time // key: jobID
 	progressUpdateTimesMu sync.RWMutex
+	// Task distribution serialization
+	taskDistMu sync.Mutex // Mutex to prevent concurrent distribution
 }
 
 // NewServer creates a new API server
@@ -78,6 +83,7 @@ func NewServer(db *database.DB, auth *authpkg.Auth, storage *storage.Storage) (*
 			WriteBufferSize: 1024,
 		},
 		runnerConns:          make(map[int64]*websocket.Conn),
+		runnerConnsWriteMu:   make(map[int64]*sync.Mutex),
 		frontendConns:        make(map[string]*websocket.Conn),
 		frontendConnsWriteMu: make(map[string]*sync.Mutex),
 		jobListConns:         make(map[int64]*websocket.Conn),
@@ -611,7 +617,7 @@ func (s *Server) recoverStuckTasks() {
 
 	go func() {
 		for range distributeTicker.C {
-			s.distributeTasksToRunners()
+			s.triggerTaskDistribution()
 		}
 	}()
 
@@ -675,7 +681,7 @@ func (s *Server) recoverStuckTasks() {
 			s.recoverTaskTimeouts()
 
 			// Distribute newly recovered tasks
-			s.distributeTasksToRunners()
+			s.triggerTaskDistribution()
 		}()
 	}
 }