The geographically-distributed multicluster computer system (CS) has been created and is being developed by Computer centre for parallel technologies of Siberian state university for telecommunications and information sciences (CPCT SibSUTIS) in conjunction with Laboratory of computer systems of A.V.Rzhanov Institute of semiconductor physics of Siberian branch of Russian Academy of Sciences (ISP SBRAS). The system composes geographically-distributed clusters and system software. Architecture of the given system is based on results of Russian scientific school on the distributed computer systems. 1. Architecture of the geographically-distributed multicluster computer system. The current system configuration unites 9 geographically-distributed cluster computer systems, seven of which are located in CPCT SibSUTIS (geographically it is the center of Novosibirsk) and the other two in ISP SBRAS (Academgorodok). The system includes a cluster, which is a part of distributed infrastructure of the program University cluster. 2. Software. It includes: standard components and original toolkit for parallel multiprogramming. Standard software components include: network operating system (GNU/Linux), tools for developing, debugging and analyzing of serial and parallel programs (MPI library: MPICH2, OpenMPI), the software for organization of system functioning (batch processing system TORQUE, scheduler MAUI), for interactions of the geographically-distributed clusters (Globus Toolkit) and for job dispatching (GridWay). The original toolkit includes: means for self-checking and self-diagnostics, a subsystem for organization of system functioning in multiprogramming modes, including: the environment for nesting of parallel programs and implementation of effective collective communications between their branches, the distributed job queue and the manager of user queries, tools for analysis of parallel programs, services for monitoring and organization of remote access to the system resources. The system is used for doing researches in the area of distributed processing of the information, debugging of tools for parallel multiprogramming and preparation of experts in the area of fault-tolerant computing technologies. The system has confirmed efficiency of architectural decisions and perspective of formation of the regional computer systems and their application for the solving of superchallenging jobs and for modeling of the modern technological processes and the natural phenomena.

Представлены архитектура и функциональная структура пространственно-распределенной мультикластерной вычислительной системы, созданной и развиваемой Центром параллельных вычислительных технологий Сибирского государственного университета телекоммуникаций и информатики совместно с Институтом физики полупроводников им. А.В. Ржанова СО РАН. Приведено описание программного обеспечения системы, включающее средства, созданные коллективом ведущей научной школы по распределенным вычислительным системам (НШ 5176.2010.9).