Development Algorithmic Model for optimization of Distributed Fault-Tolerant Web-Systems

Algebraic optimization methods for balancing work, as well as a grid model, organization of load distribution, which is the basis of a hybrid transfer mechanism are considered in this paper. The theoretical foundations, which will ensure the effective functioning of the failure hierarchy and allow to effectively calculate the load of module nodes and nodes in general are considered. The principle of operation and functionality of the grid model, which is significantly different from the usual serial and parallel models was also considered. The main difference is the ability to aggregate and share large sets of heterogeneous resources distributed among geographically separated regions. In many cases, this brings great benefits, for example, when the system requires resources that are not available within the same node, it can receive resources from other nodes connected to the grid. The grid requires a change in the concept of an application error a Web-based network service generates an error message as soon as it is impossible to execute successfully.