حصول اطمینان از ایمنی داده ها در سیستم های توزیع شده
Abstract
1- Introduction
2- Methods for ensuring data security in distributed systems
3- The decision-making process on methods of ensuring the data safety
4- Optimal placement of replicas in a distributed system.
5- Use of a reserve of replica copies and/or replica prehistory's in the network nodes
6- Application in logistics and transportation
7- Conclusion
References
Abstract
An important problem of designing distributed systems is to provide a high level of data safety. To solve this problem, two main methods could be considered. The first method is to place the replicas of data files into computer network nodes. The second method is based on using a redundancy of data files’ replicas (copies and prehistory’s of replicas) in the network nodes. The paper considers formal models of these two methods that are followed by a brief description of the decision-making process to ensure the preservation of data safety. This process is supported by means of an automated complex consisting of two software modules. These modules solve the problem of the optimal use of the two presented methods, ensuring the data safety in distributed systems. Finally, the application of the proposed approach in the design of logistics and transportation systems is discussed.
Introduction
Distributed data processing systems based on computer networks are very complex technical systems. A huge number of users communicate with these systems by means of various devices, such as workstations, laptops, tablets, smartphones, and smartwatches, etc. The work of distributed systems is based on the interaction of a large number of different software, hardware and information components. At the same time, due to the impact of various negative factors, almost all of these components can fail. Elimination of failures consequences in the distributed data processing system requires a large time and resources consuming. These costs, in turn, lead to downtime in the work of companies. As a result, the companies incur significant losses, direct and indirect. Therefore, one of the most important tasks of designing distributed systems is to ensure high system performance and high reliability of the system.