Abstract

The article solves the problem of operative selection of the redundant onboard equipment complex components configuration of the suitable in the current operating conditionаs in the interests of ensuring high fault tolerance of the complex, as well as achieving other operational and technical characteristics. The basis of the redundancy management system of the complex consists of configuration supervisors — as program subjects according to the number of its competitive configurations of heterogeneous and nonuniform equipment worked out in advance. The choice of the preferred configuration is proposed to be carried out by performing multi-level arbitration, which includes two phases of paired arbitration of computers and paired arbitration of configuration. It is proposed to include the means of both types of arbitration in each configuration supervisor, which ensures its self-sufficiency when participating in a competitive selection. The second part of the article is devoted to the computer’s arbitration for the implementation of redundancy management functions. The approach is applicable to a computing environment with many comparable computing devices and contains 2 phases. In the first phase, a preliminary selection of a competing pair of computers — as applicants for the implementation of redundancy management functions in them is carried out. In the break between the phases, the pair computers implement the procedures for pair arbitration of configurations given in the first part of the article. In the second phase, the final choice of the -computer is made, in which the supervisor who won the arbitration will be implemented. In order to achieve the maximum possible centralization of selection procedures and, as a consequence, the exclusion of " bottlenecks" in terms of reliability of places, additionally proposed: the organization of secure data exchange between computers based on distributed registry technology; the procedure of paired arbitration of computers, consisting in mutual cross-validation of dominant supervisors of a pre-allocated pair by comparing preference matrices, including information parcels of arbitration objects. A methodological example that demonstrates the features of the system functioning in the conditions of computers degradation is given. The proposed approach can be used to solve the problems of reconfiguration control of heterogeneous computing facilities of technical objects on-board equipment complexes.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call