For example,Бобцов

RELIABILITY OF MULTI-CLUSTER SYSTEMS WITH REDISTRIBUTION OF THE FLOW OF REQUESTS

Annotation

Distributed computer system under consideration consists of local clusters connected through a network; each cluster receives a separate flow of requests. The possibilities of increasing the reliability of the multi-cluster systems as a result of redistribution of requests between any pair of clusters aimed at adaptation of the system to server failures accumulation and to changes in the flow of requests. Conditions of the system performance are specified with the account for its critical dependence on servers’ overload and to excess of maximum allowable waiting time of requests in the queue.

Keywords

Articles in current issue