An approach to the total monitoring of large scale supercomputing systemsдоклад на конференции