Study On Objects Persistence And Performance Optimization Based On Relational Database

One of the most common architectures in enterprise systems today is thecombination of object-oriented applications with active relational database systems,which could lead to the mismatch between object orientation and the relational model.The object paradigm is based on software engineering principles such as coupling,cohesion and encapsulation, whereas the relational paradigm is based on mathematicalprinciples, particularly those of set theory. The mismatch comes to the problem whendesigning with object paradigm as storing with relational paradigm: with the objectparadigm you traverse objects via their relationships, with the relational paradigm youduplicate data to join the rows in tables. In the other hand, the ideal storage mechanismfor object model –object -oriented database is not mature, which is an important factorin choosing relational database in development. Therefore, it is so significant to findan approach to transfer object models to relational models and integrate of objectorientation aspects and relational database.The core of realizing the transfer is object persistence --a way that turns atransient object to a persistent object. Today a straight SQL-embedded or data accessclass solution is a solution for the mismatch, but the ways depress application'smaintainability. Once database or physical model changed, data access code inapplication should be changed correspondingly. A better solution is data access layer,which is decoupling with user interface and business logic. Every layer communicatesadjacent layer through interfaces, therefore code had changed in some layer would notinfluence other layer much.Firstly this paper analyses the characters of object model and relational model,describes object-relational mapping theory and adaptablily of mapping strategies.Secondly object persistence based on relational database and several design patterns indata access layer is discussed. Through analysis of data accessor pattern, transactionpattern ,cache pattern and each pattern's applicability, we presented a lightweightobject persistence layer framework with performance optimization and specifieddesign details.The object persistence layer framework can encapsulates database access,achieves object transparent store with cache and lazy loading. It's a lightweight andeffective database middleware for OO applications that can improve the efficiency andshorter the cycle of development owing to decoupling data access code from businesslogic. Moreover, the framework can support lots of database productions or datasources that improve transplantablity of application .So it is a best solution for objectpersistence based on relational database. This paper issues that when developing web applications developers canintegrates Struts and object persistence layer, the combination patterns can faster webapplication development and increase the quality as well as the efficiency.