首页 > 开发 > Java > 正文

Hibernate缓存详解

2024-07-13 10:03:37
字体:
来源:转载
供稿:网友

1. 什么是缓存?

数据库的缓存指的是应用程序和物理数据源之间的数据。即把物理数据源的数据复制到缓存。有了缓存,可以降低应用程序对物理数据源的访问频率,从而提高效率。缓存的介质一般是内存,也可以是硬盘。

Hibernate的缓存有三种类型:一级缓存、二级缓存和查询缓存。

2. 一级缓存

一级缓存即Session缓存,由Session自动进行管理,不需要程序进行干预。一级缓存根据对象的ID进行加载和缓存。如下面的代码:

@Override  public void testCache() {    // TODO Auto-generated method stub    Session session = sessionFactory.openSession();    Transaction tx = session.beginTransaction();     Course c = (Course) session.get(Course.class, 1);    System.out.println("Name:" + c.getName());    c = (Course) session.get(Course.class, 1);    System.out.println("Name:" + c.getName());    tx.commit();    session.close();  }

运行结果:

Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Name:计算机原理Name:计算机原理

第1次查询时生成了SQL语句,并将查询出来的对象放在一级缓存里面,第2次查询时,在一级缓存里面直接找到了这个对象,就不需要再次生成SQL语句了。

再看一个例子:

@Override  public void testCache() {    // TODO Auto-generated method stub    Session session = sessionFactory.openSession();    Transaction tx = session.beginTransaction();     Course c = (Course) session.get(Course.class, 1);    System.out.println("Name:" + c.getName());    tx.commit();    session.close();    session = sessionFactory.openSession();    tx = session.beginTransaction();     c = (Course) session.get(Course.class, 1);    System.out.println("Name:" + c.getName());    tx.commit();    session.close();  }

由于一级缓存是Session级别的缓存,所以Session关闭以后,一级缓存也就不存在了,第2次查询也要生成SQL语句:

Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Name:计算机原理Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Name:计算机原理

 3. 二级缓存

二级缓存即SessionFactory缓存,和一级缓存类似,也是根据对象的ID进行加载和缓存,区别就在于一级缓存只在Session内有效,而二级缓存在SessionFactory内有效。在访问某个ID的对象时,先到一级缓存里面去找,如果没有找到就到二级缓存里面去找。二级缓存包括EHCache,OSCache,SwarmCache和JBossCache等。这里以EHCache作为例子。

二级缓存需要程序进行管理。首先,配置Maven下载相关的Jar,在pom文件里面添加:

<dependency>       <groupId>org.hibernate</groupId>       <artifactId>hibernate-ehcache</artifactId>       <version>4.1.0.Final</version>     </dependency>    <dependency>       <groupId>net.sf.ehcache</groupId>       <artifactId>ehcache</artifactId>       <version>2.8.3</version>     </dependency>

创建EHCache配置文件ehcache.xml:

<ehcache>  <diskStore path="E:/Eclipse/MyWorkspace/Cache"/>  <defaultCache    maxElementsInMemory="10000"    eternal="true"    timeToIdleSeconds="120"    timeToLiveSeconds="120"    overflowToDisk="true"  />  <cache name="com.hzhi.course.entity.Course"    maxElementsInMemory="10000"    eternal="true"    timeToIdleSeconds="300"    timeToLiveSeconds="600"    overflowToDisk="true"  /></ehcache>

defaultCache是默认的设置,下面一个cache指明了对哪一个类进行二级缓存。里面设置了最大缓存的对象数量,是否永久有效、最大空闲秒数、最大生存秒数、内存满时是否写到硬盘,写到硬盘的路径等等。

修改需要缓存的类的hbm文件:

 <class name="com.hzhi.course.entity.Course" table="clas">    <cache usage="read-only"/>        ......  </class>

usage设置了并发访问策略,一般设置成read-only。

修改applicationContext.xml中的SessionFactory的配置,增加二级缓存的一些属性:

<!-- SessionFactory -->   <bean id="sessionFactory" class="org.springframework.orm.hibernate4.LocalSessionFactoryBean">    <property name="dataSource" >      <ref local="dataSource"/>    </property>    <!-- 配置Hibernate的属性 -->    <property name="hibernateProperties">      <props>        <prop key="hibernate.dialect">org.hibernate.dialect.MySQLDialect</prop>        <prop key="hibernate.show_sql">true</prop>        <prop key="hibernate.format_sql">true</prop>        <prop key="hibernate.connection.isolation">8</prop>        <!-- 二级缓存 -->        <prop key="hibernate.cache.use_second_level_cache">false</prop>        <prop key="hibernate.cache.region.factory_class">org.hibernate.cache.ehcache.EhCacheRegionFactory</prop>                <prop key="hibernate.cache.provider_configuration_file_resource_path">WEB-INF/ehcache.xml</prop>              </props>    </property>     ......   </bean>

运行下面的例子:

@Override  public void testCache() {    // TODO Auto-generated method stub    Session session = sessionFactory.openSession();    Transaction tx = session.beginTransaction();     Course c = (Course) session.get(Course.class, 1);    System.out.println("Name:" + c.getName());    tx.commit();    session.close();    session = sessionFactory.openSession();    tx = session.beginTransaction();     c = (Course) session.get(Course.class, 1);    System.out.println("Name:" + c.getName());    tx.commit();    session.close();  }

结果:

Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Name:计算机原理Name:计算机原理

虽然关闭了Session,但是二级缓存仍然存在,所以只生成了一次SQL语句。

下面的例子:

@Override  public void testCache() {    // TODO Auto-generated method stub    Session session = sessionFactory.openSession();    Transaction tx = session.beginTransaction();     Query query = session.createQuery("from Course");     Iterator iter = query.iterate();     while(iter.hasNext()){         System.out.println(((Course)iter.next()).getName());     }    tx.commit();    session.close();        session = sessionFactory.openSession();    tx = session.beginTransaction();     query = session.createQuery("from Course");     iter = query.iterate();     while(iter.hasNext()){         System.out.println(((Course)iter.next()).getName());     }    tx.commit();    session.close();  }

结果:

Hibernate:   select    course0_.ID as col_0_0_   from    clas course0_Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?计算机原理Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?计算机网络Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?数据库原理Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?C语言Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?大学英语AHibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?JavaHibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?LinuxHibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?高等数学Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?语文Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?大学物理Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?软件工程Hibernate:   select    course0_.ID as col_0_0_   from    clas course0_计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程

当使用Query的list()方法时,只生成一次SQL语句查询出所有的对象,使用iterate()方法时,会先得到所有对象的ID,然后根据每个ID生成一次SQL语句查询。第二个Session里面使用的也是iterate()方法,首先生成一次SQL语句,得到ID,然后根据ID查找对象,由于开启了二级缓存,在二级缓存里面找到了对象,所以就直接输出了,并没有再根据每个ID生成SQL语句。

不论是一级缓存还是二级缓存,都只能缓存对象,不能缓存属性的值。下面的例子:

@Override  public void testCache() {    // TODO Auto-generated method stub    Session session = sessionFactory.openSession();    Transaction tx = session.beginTransaction();     Query query = session.createQuery("select c.name from Course c");      List<String> names = query.list();     for(Iterator iter = names.iterator(); iter.hasNext();){       String name = (String) iter.next();       System.out.println(name);     }     System.out.println("----------");     query = session.createQuery("select c.name from Course c");      names = query.list();     for(Iterator iter = names.iterator(); iter.hasNext();){       String name = (String) iter.next();       System.out.println(name);     }     System.out.println("----------");     tx.commit();    session.close();  }

运行结果:

Hibernate:   select    course0_.NAME as col_0_0_   from    clas course0_计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------Hibernate:   select    course0_.NAME as col_0_0_   from    clas course0_计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------

虽然开启了二级缓存,但是查询的结果不是对象,是属性,所以并没有缓存,第2次查询仍然生成了查询语句。要解决这个问题,就需要查询缓存。

3. 查询缓存

在配置了二级缓存的基础上,可以设置查询缓存,在SessionFactory的设置里面加上一行:

<prop key="hibernate.cache.use_query_cache">true</prop>

即打开了查询缓存。查询缓存也是SessionFactory级别的缓存,在整个SessionFactory里面都是有效的。

关闭二级缓存,运行下面的例子,在Query后面添加setCacheable(true)打开查询缓存:

@Override  public void testCache() {    // TODO Auto-generated method stub    Session session = sessionFactory.openSession();    Transaction tx = session.beginTransaction();     Query query = session.createQuery("select c.name from Course c");     query.setCacheable(true);     List<String> names = query.list();     for(Iterator iter = names.iterator(); iter.hasNext();){       String name = (String) iter.next();       System.out.println(name);     }     System.out.println("----------");     query = session.createQuery("select c.name from Course c");     query.setCacheable(true);     names = query.list();     for(Iterator iter = names.iterator(); iter.hasNext();){       String name = (String) iter.next();       System.out.println(name);     }     System.out.println("----------");     tx.commit();    session.close();  }

结果:

Hibernate:   select    course0_.NAME as col_0_0_   from    clas course0_计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------

由于两次查询的HQL语句是一致的,所以只生成一次SQL语句。但是如果把第二次查询改一下:

System.out.println("----------");     query = session.createQuery("select c.name from Course c where c.id > 5");     query.setCacheable(true);     names = query.list();     for(Iterator iter = names.iterator(); iter.hasNext();){       String name = (String) iter.next();       System.out.println(name);     }     System.out.println("----------");

结果:

Hibernate:   select    course0_.NAME as col_0_0_   from    clas course0_计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------Hibernate:   select    course0_.NAME as col_0_0_   from    clas course0_   where    course0_.ID>5大学英语AJavaLinux高等数学语文大学物理软件工程----------

由于HQL语句变了,所以第二次也生成了SQL语句。

查询缓存可以缓存属性,也可以缓存对象,但是当缓存对象时,只缓存对象的ID,不会缓存整个对象。下面的例子:

@Override  public void testCache() {    // TODO Auto-generated method stub    Session session = sessionFactory.openSession();    Transaction tx = session.beginTransaction();     Query query = session.createQuery("from Course");    query.setCacheable(true);    List<Course> list = query.list();    for (int i=0; i<list.size(); i++){      System.out.println(list.get(i).getName());     }    System.out.println("----------");     tx.commit();    session.close();    session = sessionFactory.openSession();    tx = session.beginTransaction();        query = session.createQuery("from Course");     query.setCacheable(true);    list = query.list();    for (int i=0; i<list.size(); i++){      System.out.println(list.get(i).getName());     }    System.out.println("----------");    tx.commit();    session.close();  }

 结果:

 Hibernate:   select    course0_.ID as ID0_,    course0_.NAME as NAME0_,    course0_.COMMENT as COMMENT0_   from    clas course0_计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?Hibernate:   select    course0_.ID as ID0_0_,    course0_.NAME as NAME0_0_,    course0_.COMMENT as COMMENT0_0_   from    clas course0_   where    course0_.ID=?计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------

由于开了查询缓存,没有开二级缓存,虽然使用的是list()方法一次查询出了所有的对象,但是查询缓存只缓存了对象ID,没有缓存整个对象。所以在第2个Session里面"from Course"这个HQL由于和前面的相同,并没有生成SQL语句,但是由于没有开二级缓存,没有缓存整个对象,只能根据每个ID去生成一次SQL语句。虽然两次用的都是list()方法,但是第一次是生成SQL语句去一次查询出所有的对象,而第二次是根据查询缓存里面的ID一个一个的生成SQL语句。

如果同时打开查询缓存和二级缓存,第2个Session里面就不用再根据ID去生成SQL语句了:

Hibernate:   select    course0_.ID as ID0_,    course0_.NAME as NAME0_,    course0_.COMMENT as COMMENT0_   from    clas course0_计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------计算机原理计算机网络数据库原理C语言大学英语AJavaLinux高等数学语文大学物理软件工程----------

以上就是本文的全部内容,希望本文的内容对大家的学习或者工作能带来一定的帮助,同时也希望多多支持VeVb武林网!


注:相关教程知识阅读请移步到JAVA教程频道。
发表评论 共有条评论
用户名: 密码:
验证码: 匿名发表