1.简介
MyBatis 是支持普通SQL 查询,存储过程和高级映射的优秀持久层框架。MyBatis 消除了几乎所有的JDBC 代码和参数的手工设置以及结果集的检索。
Ehcache 是现在最流行的纯Java开源缓存框架,配置简单、结构清晰、功能强大,最初知道它,是从Hibernate的缓存开始的。
2. 准备工作
下载mybatis相关包与ehcache相关包
ehcache-core-2.4.4.jar
mybatis-ehcache-1.0.0.jar
slf4j-api-1.6.1.jar
slf4j-log4j12-1.6.2.jar
3. 配置步骤:
用hsqldb作为数据库,使用mybatis自定义缓存。
数据库建表语句:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
create table category ( catid varchar(10) not null, name varchar(80) null, descn varchar(255) null, constraint pk_category primary key (catid));create table product ( productid varchar(10) not null, category varchar(10) not null, name varchar(80) null, descn varchar(255) null, constraint pk_product primary key (productid), constraint fk_product_1 foreign key (category) references category (catid));create index productCat on product (category);create index productName on product (name); |
insert数据的语句:
|
1
2
3
4
|
INSERT INTO category VALUES ('FISH','Fish','<image src="../images/fish_icon.gif"><font size="5" color="blue"> Fish</font>');INSERT INTO category VALUES ('DOGS','Dogs','<image src="../images/dogs_icon.gif"><font size="5" color="blue"> Dogs</font>');INSERT INTO product VALUES ('FI-SW-01','FISH','Angelfish','<image src="../images/fish1.gif">Salt Water fish from Australia');INSERT INTO product VALUES ('FI-SW-02','FISH','Tiger Shark','<image src="../images/fish4.gif">Salt Water fish from Australia'); |
4. cache配置步骤:
(1).在classpath下配置ehcache.xml。这个是ehcache的默认配置文件。
|
1
2
3
4
5
6
|
<?xml version="1.0" encoding="UTF-8"?> <ehcache xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:noNamespaceSchemaLocation="../bin/ehcache.xsd"> <defaultCache overflowToDisk="true" eternal="false" maxElementsInMemory="1"/> <diskStore path="D:/cache" /> </ehcache> |
(2).在需要的Mapper.xml中配置cache。
上面的配置是全局的cache,在Mapper.xml中可以根据自己的需要,对这个Mapper中进行cache的配置,可以配置某一条sql语句不进行cache。
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
<cache type="org.mybatis.caches.ehcache.LoggingEhcache"/> //最普通的设置,沿用全局设置<cache type="org.mybatis.caches.ehcache.LoggingEhcache" > <property name="timeToIdleSeconds" value="3600"/><!--1 hour--> <property name="timeToLiveSeconds" value="3600"/><!--1 hour--> <property name="maxEntriesLocalHeap" value="1000"/> <property name="maxEntriesLocalDisk" value="10000000"/> <property name="memoryStoreEvictionPolicy" value="LRU"/><cache><!--配置这个mapper使用LRU替换策略。(个人比较赞同这种配置,因为每个表的数据都不一样,有一些需要经常更新,有得可能某几个字段需要经常做连接,使用一样的cache不太合适)--> |
mybatis默认是启用cache的,所以对于某一条不想被cache的sql需要把useCache="false"
|
1
|
<select id="getCategory" parameterType="string" resultType="Category" useCache="false"> |
5. 测试:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
|
private ApplicationContext application;private CategoryMapper categoryMapper;private ProductMapper productMapper;private String[] categoryId = { "FISH", "DOGS", "REPTILES", "CATS", "BIRDS" };@Beforepublic void initSpring() { application = new FileSystemXmlApplicationContext( "resource/applicationContext.xml"); categoryMapper = application.getBean(CategoryMapper.class); productMapper = application.getBean(ProductMapper.class);}@Testpublic void testSelect() { // the first time long begin = System.nanoTime(); categoryMapper.getCategory(categoryId[0]); long end = System.nanoTime() - begin; print("count :" + end); // the second time begin = System.nanoTime(); categoryMapper.getCategory(categoryId[0]); end = System.nanoTime() - begin; print("count :" + end); // the third time begin = System.nanoTime(); categoryMapper.getCategory(categoryId[0]); end = System.nanoTime() - begin; print("count :" + end); //}@Testpublic void testInsert() { // the second time long begin = System.nanoTime(); // Product p1 = productMapper.getProduct("FI-SW-01"); long end = System.nanoTime() - begin; print("count :" + end); print("Category :"+p1.getCategoryId()); Map<String, String> parame = new HashMap<String, String>(); parame.put("categoryId", "DOGS"); parame.put("productId", "FI-SW-01"); begin = System.nanoTime(); productMapper.updateProductById(parame); end = System.nanoTime() - begin; print("count :"+end); begin = System.nanoTime(); Product p2 = productMapper.getProduct("FI-SW-01"); end = System.nanoTime() - begin; print("count :"+end); print("Category :"+p2.getCategoryId()); } |
运行的结果
log4j debug开启:
16 09:42:16,447 org.apache.ibatis.logging.slf4j.Slf4jImpl: Cache Hit Ratio [org.mybatis.jpetstore.persistence.CategoryMapper]: 0.0
DEBUG 2013-05-16 09:42:16,501 org.apache.ibatis.logging.slf4j.Slf4jImpl: ==> Executing: SELECT CATID AS categoryId, NAME, DESCN AS description FROM CATEGORY WHERE CATID = ?
DEBUG 2013-05-16 09:42:16,502 org.apache.ibatis.logging.slf4j.Slf4jImpl: ==> Parameters: FISH(String)
DEBUG 2013-05-16 09:42:16,542 org.apache.ibatis.logging.slf4j.Slf4jImpl: <== Columns: CATEGORYID, NAME, DESCRIPTION
DEBUG 2013-05-16 09:42:16,542 org.apache.ibatis.logging.slf4j.Slf4jImpl: <== Row: FISH, Fish, <image src="../images/fish_icon.gif"><font size="5" color="blue"> Fish</font>
count :204671120
Cache Hit Ratio [org.mybatis.jpetstore.persistence.CategoryMapper]: 0.5
count :3737320
Cache Hit Ratio [org.mybatis.jpetstore.persistence.CategoryMapper]: 0.6666666666666666
count :2349519
这里可以很清晰的看到,第一次取数据的时候,mybatis运行了sql语句,并且得到了返回的数据,因为是第一次,所以 Cache Hit Ratio 是0.0,当第二次的时候,因为cache已经存在查询的数据集,因此,mybatis没有发起查询,直接得到了数据。两次的时间差了100倍。
log4j INFO,关掉debug得到的查询速度快了一点
count :126226400
count :1803960
count :772080inset:语句的运行。首先查一遍,让cache里面有数据,然后update里面的数据,再取出来。从所耗费的时间来看,update只是在cache进行了,只有等cache过期了,数据才会写入数据库。
count :121264080
Category :FISH
count :4028440
count :1509640
Category :DOGS6.总结思考
查看了一下硬盘cache的文件夹,发现mybatis对于每个Mapper都有自己独立的cache文件,查看mybatis-ehcache的实现:
private static final CacheManager CACHE_MANAGER = CacheManager.create();
每一个Mapper都对应的有自己的CacheManager。
虽然,通过log可以看到cache hit radio 但是,我需要像ehcache整合到spring那样子可以获得每个由mybatis代理运行的cache对象。查看了很久的代码,发现里面的代理对象都是私有的,并不能被用户调用,所以现在还没有找到细粒化操作mybatis cache的方法。
< ehcache xmlns:xsi = "http://www.w3.org/2001/XMLSchema-instance" xsi:noNamespaceSchemaLocation = "ehcache.xsd" >
< diskStore path = "java.io.tmpdir" />
< defaultCache
maxElementsInMemory = "10000"
maxElementsOnDisk = "0"
eternal = "true"
overflowToDisk = "true"
diskPersistent = "false"
timeToIdleSeconds = "0"
timeToLiveSeconds = "0"
diskSpoolBufferSizeMB = "50"
diskExpiryThreadIntervalSeconds = "120"
memoryStoreEvictionPolicy = "LFU"
/>
< cache name = "myCache"
maxElementsInMemory = "100"
maxElementsOnDisk = "0"
eternal = "false"
overflowToDisk = "false"
diskPersistent = "false"
timeToIdleSeconds = "120"
timeToLiveSeconds = "120"
diskSpoolBufferSizeMB = "50"
diskExpiryThreadIntervalSeconds = "120"
memoryStoreEvictionPolicy = "FIFO"
/>
</ ehcache >
< diskStore path = "java.io.tmpdir" />
< defaultCache
maxElementsInMemory = "10000"
maxElementsOnDisk = "0"
eternal = "true"
overflowToDisk = "true"
diskPersistent = "false"
timeToIdleSeconds = "0"
timeToLiveSeconds = "0"
diskSpoolBufferSizeMB = "50"
diskExpiryThreadIntervalSeconds = "120"
memoryStoreEvictionPolicy = "LFU"
/>
< cache name = "myCache"
maxElementsInMemory = "100"
maxElementsOnDisk = "0"
eternal = "false"
overflowToDisk = "false"
diskPersistent = "false"
timeToIdleSeconds = "120"
timeToLiveSeconds = "120"
diskSpoolBufferSizeMB = "50"
diskExpiryThreadIntervalSeconds = "120"
memoryStoreEvictionPolicy = "FIFO"
/>
</ ehcache >
diskStore :指定数据存储位置,可指定磁盘中的文件夹位置
defaultCache : 默认的管理策略
以下属性是必须的:
- name: Cache的名称,必须是唯一的(ehcache会把这个cache放到HashMap里)。
- maxElementsInMemory: 在内存中缓存的element的最大数目。
- maxElementsOnDisk: 在磁盘上缓存的element的最大数目,默认值为0,表示不限制。
- eternal: 设定缓存的elements是否永远不过期。如果为true,则缓存的数据始终有效,如果为false那么还要根据timeToIdleSeconds,timeToLiveSeconds判断。
- overflowToDisk: 如果内存中数据超过内存限制,是否要缓存到磁盘上。
以下属性是可选的:
- timeToIdleSeconds: 对象空闲时间,指对象在多长时间没有被访问就会失效。只对eternal为false的有效。默认值0,表示一直可以访问。
- timeToLiveSeconds: 对象存活时间,指对象从创建到失效所需要的时间。只对eternal为false的有效。默认值0,表示一直可以访问。
- diskPersistent: 是否在磁盘上持久化。指重启jvm后,数据是否有效。默认为false。
- diskExpiryThreadIntervalSeconds: 对象检测线程运行时间间隔。标识对象状态的线程多长时间运行一次。
- diskSpoolBufferSizeMB: DiskStore使用的磁盘大小,默认值30MB。每个cache使用各自的DiskStore。
- memoryStoreEvictionPolicy: 如果内存中数据超过内存限制,向磁盘缓存时的策略。默认值LRU,可选FIFO、LFU。
缓存的3 种清空策略 :
FIFO ,first in first out (先进先出).
LFU , Less Frequently Used (最少使用).意思是一直以来最少被使用的。缓存的元素有一个hit 属性,hit 值最小的将会被清出缓存。
LRU ,Least Recently Used(最近最少使用). (ehcache 默认值).缓存的元素有一个时间戳,当缓存容量满了,而又需要腾出地方来缓存新的元素的时候,那么现有缓存元素中时间戳离当前时间最远的元素将被清出缓存。
FIFO ,first in first out (先进先出).
LFU , Less Frequently Used (最少使用).意思是一直以来最少被使用的。缓存的元素有一个hit 属性,hit 值最小的将会被清出缓存。
LRU ,Least Recently Used(最近最少使用). (ehcache 默认值).缓存的元素有一个时间戳,当缓存容量满了,而又需要腾出地方来缓存新的元素的时候,那么现有缓存元素中时间戳离当前时间最远的元素将被清出缓存。