| The rapid development of cloud computing technology has made cloud databases popular.However,the data stored in cloud databases is facing the risk of leakage since the separation of the use rights and ownership of the cloud data.Encrypting the data by encrypted databases before storing it in the cloud databases is one of the feasible solutions of protecting private data.However,the system cannot guarantee availability while taking into security due to the inefficient processing of complex statements by the encrypted database system as well as the lack of an encrypted resource management in a highly concurrent environment,so it cannot be widely used.Therefore,this thesis optimizes the system of encrypted databases,which is of great significance to the promotion and application of encrypted database.The performance test of the encrypted database prototype system in the early stage found that the challenge of improving efficiency is that the processing efficiency of the complex query statements involved in the actual application is low first,because the internal parsing component of the system generate large-scale and complex structure AST.The construction and traversal of AST is very time-consuming.Furthermore,encrypted databases cannot effectively respond to user requests in high concurrent environments.Therefore,this thesis optimize the AST generated while the encrypted database processing complex SQL statements and an effective resource scheduling for the encrypted resources of the system is proposed.The main tasks of this thesis are as follows:1)A method of constructing Clamshell graph is proposed to optimize the abstract syntax tree structure generated by the parsing module when processing complex queries.It will reduce the redundant nodes in AST.Meanwhile,9 types of optimizing module is designed base on the different types of complex SQL statement is classified.On this basis,the syntax tree in the parsing module.On this basis,the syntax tree in the parsing module is specifically optimized,and an encrypted database parsing module that could efficiently processes complex SQL statements is realized.2)An effective resource scheduling scheme is proposed for the encrypted database system,which can effectively deal with the situation that multiple users concurrently request to call the encryption service module.First,an effective key distribution scheme under high concurrency scenarios is designed,then an efficient encryption resource scheduling scheme is proposed.By setting priority numbers for the SQL statements that are passed into the encryption service module.To improve the utilization of encryption resources,those resources are allocated to the SQL statement according to the priority number,and thus the overall throughput and performance of the system is improved.3)Performance tests were carried out in the prototype system to verify the availability of the optimization solution.The result show that the encryption database can significantly improve the processing efficiency of complex SQL statements.Furthermore,encryption resources could be reasonably scheduled and hence effectively cope with the scenario of multi-user concurrent request data in the cloud environment.The performance optimization scheme under high concurrent environment is designed for the encrypted database,and the parsing module and resource management module in the system are re-implemented.Tests were carried out based on the optimized encrypted database system,the results show that the tasks processing efficiency and resource utilization rate have been significantly improved in the high concurrent environment,and the expected goal has been reached. |
