The Method for Ensuring the Survivability of Distributed Computing in Heterogeneous Computer Systems

Nowadays specialists in distributed computing try to reduce execution time for complex calculations. One of the tasks about increasing the efficiency of distributed computing is - ensure survivability, which consists in the fastest possible restoration of the distributed program system in the event of the failure of some hardware resources. This article is devoted to the study the issues of ensuring the functional reliability of software. The rollback process formalization for survivability of a distributed computing are presented. The rules of resource management are formulated. Method of providing survivability by memory dump of software components was developed. This method can be implemented as a separate middleware or by embedding it in distributed software, for example, as a library of functions.