Research On System Fault Tolerance Strategy Based On Service Redundancy In Cloud Environment

With the continuous development of cloud computing technology,Service-Based System SBS is growing rapidly,which combines existing services in the form of business processes to provide services to users.However,this architecture also brings more complex problems to the fault tolerance of the system.Especially in the dynamic cloud environment,the component services of SBS may experience quality degradation caused by various runtime anomalies,such as virtual machine faults,software exceptions,etc.Timely and cost-effectively handling runtime exceptions of multi-tenant SBS is critical to ensure the quality of the SBS.The main works of this thesis are as follows:(1)A system fault tolerance strategy based on service redundancy is proposed.In order to formulate a reasonable redundancy strategy,we need to sort the component services critically,so that the more critical component services can get better redundancy scheme.Due to the dynamic nature of the cloud environment,the value of the QoS of component services may be variable in different time periods.This paper uses historical data to predict the future QoS of component services when identifying critical component services.In this article,we mathematically model historical data use a normal distribution model to organize the data,and use the characteristics of normal distribution to calculate criticality for component services.In addition,a cost-effective strategy for making redundancy schemes is proposed.It is an NP problem to select the best redundancy scheme among a large number of candidate services.Then a constrained optimization problem model based on criticality of component service is established to formulate a cost-effective SBS fault-tolerant strategy.The fault-tolerant strategy assigns redundancy schemes to key component services,and use linear programming to find the best solution to the constraint optimization problem model.(2)A fault-tolerant strategy for multi-tenant service-based system is proposed.In real life,in order to maximize the use of cloud resources,system development engineers often use multi-tenant technology to build a multi-tenant system.This also brings more problems to the research of fault-tolerant strategies of the system.It is also necessary to consider the characteristics of multi-tenancy when formulating a fault-tolerant strategy.As different tenants may have different orders of users,this article first divides the multi-tenants into groups according to the agreement signed between the system development company and each tenant.In order to ensure that the end-users enjoy the benefits of fault-tolerant strategy to a great extent,tenants of different levels have different weights when performing critical calculations on component services.meanwhile,when formulating redundancy schemes,the influence of multi-tenant preference on different service quality will be fully considered.