Research And Implementation Of Key Technology Of IVCE Platform Monitoring System

Currently,each company has more data and more transactions to deal with,prompting all companies to cloud the business.The IVCE platform is a platform developed to test the status of each operator and provinces in the daily nationwide network,In order to ensure the continuous and effective test results,we must ensure that the platform can be continuous operation,failure also requires a short period of recovery,but the platform of the virtual resource performance is unstable,often there will be dropped,and may be limited Speed,and the confidential data stored therein is not encrypted and saved,developers,as a developer,must be able to overcome these difficulties.First of all,according to these requirements,this paper studies the virtual resource performance prediction algorithm based on BP neural network algorithm to provide platform managers with the performance reference of future virtual resources.Because the BP algorithm has the problem of slow convergence speed and easy to fall into the minimum value,adding the Bayesian regularization theory and the attitude and learning rate parameters to form the Trainbr-BP neural network algorithm.After the new data is collected,the neural network is reversely input again to adjust The neuron's weights can reach 99.086%and 99.99015%under the steady state,and the convergence steps are only 1/10 of the original ones.Secondly,this paper improves the Cryptography algorithm of CryptDB,applies the modular de-encryption algorithm,adds the offset to the output result to improve the data security.In the stress test and the use of different index types,In addition,there is no obvious decrease in operational efficiency with the increase of security.In addition,the binary search process in MOPE algorithm is further improved,and the plaintext space is fitted to This paper not only theoretically analyzes that the improved scheme reduces the 'average search length from log2(n + 1)-1 to 1?log2(n + 1)-1,and the design experiment proves that in different concurrent The number of sentences and different types of index database operating efficiency increased by 5.99%and 5.97%,in practice the application is also significant effect.Finally,it studies the monitoring system of the health status of IVCE platform,develops monitoring client,script and front interface to collect information,condition display and dispatching operations on all virtual resources and servers in the platform.CryptDB to Mycat secure data storage module to form a hot standby high availability database module to improve data security and database robustness and the use of JSON communication format for unified regulation,the formation of thismonitoring system.Throughout the whole article,this paper analyzes the problems of IVCE platform in detail,lists the demand according to the analysis,respectively uses BP neural network to research the resource performance prediction algorithm,and studies and improves the security and search efficiency of the order-preserving encryption algorithm,Design and implementation of the monitoring system of the various modules,the final design experiment to verify the effectiveness and availability of the system.