Refactor job status handling to prevent race conditions

- Removed redundant error handling in handleListJobTasks.
- Introduced per-job mutexes in Manager to serialize updateJobStatusFromTasks calls, ensuring thread safety during concurrent task completions.
- Added methods to manage job status update mutexes, including creation and cleanup after job completion or failure.
- Improved error handling in handleGetJobStatusForRunner by consolidating error checks.

internal/manager/jobs.go

@@ -3024,9 +3024,6 @@ func (s *Manager) handleListJobTasks(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 	defer rows.Close()
-	if err != nil {
-		total = -1
-	}
 
 	tasks := []types.Task{}
 	for rows.Next() {