首先从使用springboot+redis碰到的一个问题说起。在前几篇文章中介绍了用SpringBoot+redis构建了一个个人博客。在刚开始远行的时候发现发了几个请求操作了几次redis之后，后面的就被阻塞了，请求一直在等待返回，我们重现一下问题。
[注意] 该问题只会出现在springboot 2.0之前的版本;2.0之后springboot连接Redis改成了lettuce，并重新实现，问题已经不存在
打开Template的事务支持
POM 配置：
<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
<modelVersion>4.0.0</modelVersion>
<groupId>com.github.springboot</groupId>
<artifactId>redis-tx-demo</artifactId>
<version>0.0.1-SNAPSHOT</version>
<packaging>jar</packaging>
<name>SpringBoot redis TX demo</name>
<description>Demo project for Spring Boot with Redis transaction</description>
<parent>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-parent</artifactId>
<version>1.5.2.RELEASE</version>
<relativePath/> <!-- lookup parent from repository -->
</parent>
<properties>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
<java.version>1.8</java.version>
</properties>
<dependencies>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-data-redis</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-data-rest</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-test</artifactId>
<scope>test</scope>
</dependency>
</dependencies>
<build>
<plugins>
<plugin>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-maven-plugin</artifactId>
</plugin>
</plugins>
</build>
</project>
Redis configuration (EnbaleTransactionSupport设为true):
@Configuration
public class RedisConfiguration {
@Bean
public StringRedisTemplate redisTemplate(RedisConnectionFactory redisConnectionFactory) {
StringRedisTemplate template = new StringRedisTemplate();
template.setConnectionFactory(redisConnectionFactory);
template.setEnableTransactionSupport(true); //打开事务支持
return template;
}
}
Controller就是简单的set一个key到redis:
@RestController
public class DemoController {
private StringRedisTemplate template;
public DemoController(StringRedisTemplate template) {
this.template = template;
}
@GetMapping("/put")
public String redisSet() {
int i = (int)(Math.random() * 100);
template.opsForValue().set("key"+i, "value"+i, 300, TimeUnit.SECONDS);
return "success "+"key"+i;
}
}
启动后，我们使用RestClient发送请求http://localhost:8080/put，发送8次之后就会发现没有返回了。这个时候我们查看redis的链接数，发现已经超过8个，springboot对于jedis连接池默认的最大活跃连接数是8，所以看出来是连接池被耗光了。
127.0.0.1:6379> info clients
# Clients
connected_clients:9
client_longest_output_list:0
client_biggest_input_buf:0
blocked_clients:0
127.0.0.1:6379>
还有查看程序的日志可以发现，RedisConnectionUtils只有Opening RedisConnection而没有close。
2018-08-11 11:00:48.889 [DEBUG][http-nio-8080-exec-8]:o.s.data.redis.core.RedisConnectionUtils [doGetConnection:126] Opening RedisConnection
2018-08-11 11:00:50.169 [DEBUG][http-nio-8080-exec-8]:o.s.w.s.m.m.a.RequestResponseBodyMethodProcessor [writeWithMessageConverters:249] Written as "text/plain" using [org.springframework.http.converter.StringHttpMessageConverter@766a49c7]
2018-08-11 11:00:50.170 [DEBUG][http-nio-8080-exec-8]:org.springframework.web.servlet.DispatcherServlet [processDispatchResult:1044] Null ModelAndView returned to DispatcherServlet with name 'dispatcherServlet': assuming HandlerAdapter completed request handling
2018-08-11 11:00:50.170 [DEBUG][http-nio-8080-exec-8]:org.springframework.web.servlet.DispatcherServlet [processRequest:1000] Successfully completed request
2018-08-11 11:00:50.170 [DEBUG][http-nio-8080-exec-8]:o.s.boot.web.filter.OrderedRequestContextFilter [doFilterInternal:104] Cleared thread-bound request context: org.apache.catalina.connector.RequestFacade@c03b2d8
2018-08-11 11:00:53.854 [DEBUG][http-nio-8080-exec-9]:o.s.boot.web.filter.OrderedRequestContextFilter [initContextHolders:114] Bound request context to thread: org.apache.catalina.connector.RequestFacade@c03b2d8
2018-08-11 11:00:53.856 [DEBUG][http-nio-8080-exec-9]:org.springframework.web.servlet.DispatcherServlet [doService:865] DispatcherServlet with name 'dispatcherServlet' processing GET request for [/put]
2018-08-11 11:00:53.857 [DEBUG][http-nio-8080-exec-9]:o.s.w.s.m.m.a.RequestMappingHandlerMapping [getHandlerInternal:310] Looking up handler method for path /put
2018-08-11 11:00:53.857 [DEBUG][http-nio-8080-exec-9]:o.s.w.s.m.m.a.RequestMappingHandlerMapping [getHandlerInternal:317] Returning handler method [public java.lang.String com.github.springboot.demo.DemoController.redisSet()]
2018-08-11 11:00:53.858 [DEBUG][http-nio-8080-exec-9]:o.s.b.factory.support.DefaultListableBeanFactory [doGetBean:251] Returning cached instance of singleton bean 'demoController'
2018-08-11 11:00:53.858 [DEBUG][http-nio-8080-exec-9]:org.springframework.web.servlet.DispatcherServlet [doDispatch:951] Last-Modified value for [/put] is: -1
2018-08-11 11:00:53.861 [DEBUG][http-nio-8080-exec-9]:o.s.data.redis.core.RedisConnectionUtils [doGetConnection:126] Opening RedisConnection
关闭template的事务支持
接下来我们修改一下RedisConfiguration的配置，不启用事务管理，
@Bean
public StringRedisTemplate redisTemplate(RedisConnectionFactory redisConnectionFactory) {
StringRedisTemplate template = new StringRedisTemplate();
template.setConnectionFactory(redisConnectionFactory);
// template.setEnableTransactionSupport(true); //禁用事务支持
return template;
}
重新测试一下，发现是正常的，redis的client链接数一直保持在2。程序日志里的也可以看到Redis Connection关闭的日志。
2018-08-11 15:55:19.975 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [doGetConnection:126] Opening RedisConnection
2018-08-11 15:55:20.029 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [releaseConnection:210] Closing Redis Connection
2018-08-11 15:55:20.056 [DEBUG][http-nio-8080-exec-1]:o.s.w.s.m.m.a.RequestResponseBodyMethodProcessor [writeWithMessageConverters:249] Written as "text/plain" using [org.springframework.http.converter.StringHttpMessageConverter@51ab1ee3]
也就是说，如果我们把事务的支持打开，spring在每次操作之后是不会主动关闭连接的。我们去RedisTemplate的源码中找下原因。
public ValueOperations<K, V> opsForValue() {
if (valueOps == null) {
valueOps = new DefaultValueOperations<K, V>(this);
}
return valueOps;
}
可以发现template.opsForValue().set()操作最终是调用的DefaultValueOperations中的set()方法，继续跟进去最终调用的RedisTemplate中的execute(RedisCallback<T> action, boolean exposeConnection, boolean pipeline)方法。
public <T> T execute(RedisCallback<T> action, boolean exposeConnection, boolean pipeline) {
Assert.isTrue(initialized, "template not initialized; call afterPropertiesSet() before using it");
Assert.notNull(action, "Callback object must not be null");
RedisConnectionFactory factory = getConnectionFactory();
RedisConnection conn = null;
try {
if (enableTransactionSupport) {
// only bind resources in case of potential transaction synchronization
conn = RedisConnectionUtils.bindConnection(factory, enableTransactionSupport);
} else {
conn = RedisConnectionUtils.getConnection(factory);
}
boolean existingConnection = TransactionSynchronizationManager.hasResource(factory);
RedisConnection connToUse = preProcessConnection(conn, existingConnection);
boolean pipelineStatus = connToUse.isPipelined();
if (pipeline && !pipelineStatus) {
connToUse.openPipeline();
}
RedisConnection connToExpose = (exposeConnection ? connToUse : createRedisConnectionProxy(connToUse));
T result = action.doInRedis(connToExpose);
// close pipeline
if (pipeline && !pipelineStatus) {
connToUse.closePipeline();
}
// TODO: any other connection processing?
return postProcessResult(result, connToUse, existingConnection);
} finally {
RedisConnectionUtils.releaseConnection(conn, factory);
}
}
可以看到获取连接的操作也针对打开事务支持的template有特殊的处理逻辑。这里我们先跳过，先看看最终肯定会走到的RedisConnectionUtils.releaseConnection(conn, factory)这一步。
/**
* Closes the given connection, created via the given factory if not managed externally (i.e. not bound to the
* thread).
*
* @param conn the Redis connection to close
* @param factory the Redis factory that the connection was created with
*/
public static void releaseConnection(RedisConnection conn, RedisConnectionFactory factory) {
if (conn == null) {
return;
}
RedisConnectionHolder connHolder = (RedisConnectionHolder) TransactionSynchronizationManager.getResource(factory);
if (connHolder != null && connHolder.isTransactionSyncronisationActive()) {
if (log.isDebugEnabled()) {
log.debug("Redis Connection will be closed when transaction finished.");
}
return;
}
// release transactional/read-only and non-transactional/non-bound connections.
// transactional connections for read-only transactions get no synchronizer registered
if (isConnectionTransactional(conn, factory)
&& TransactionSynchronizationManager.isCurrentTransactionReadOnly()) {
unbindConnection(factory);
} else if (!isConnectionTransactional(conn, factory)) {
if (log.isDebugEnabled()) {
log.debug("Closing Redis Connection");
}
conn.close();
}
}
可以看到针对打开事务支持的template,只是解绑了连接，根本没有做close的操作。关于什么是解绑，其实这个方法的注释中已经说的比较清楚了，对于开启了事务的Template，由于已经绑定了线程中连接，所以这里是不会关闭的，只是做了解绑的操作。
到这里原因就很清楚了，就是只要template开启了事务支持，spring就认为只要使用这个template就会包含在事务当中，因为一个事务中的操作必须在同一个连接中完成，所以在每次get/set之后，template是不会关闭链接的，因为它不知道事务有没有结束。
使用@Transanctional注解支持Redis事务
既然RedisTemlate在setEnableTransactionSupport会造成连接不关闭，那怎么样才能正常关闭呢？我们将事务支持开关和@Transanctional结合起来用看看会怎么样。
spring中要使用@Transanctional首先要配transactionManager，但是spring没有专门针对Redis的事务管理器实现，而是所有调用RedisTemplate的方法最终都会调用到RedisConnctionUtils这个类的方法上面，在这个类里面会判断是不是进入到事务里面，也就是说Redis的事务管理的功能是由RedisConnctionUtils内部实现的。
根据官方文档，我只想用Redis事务，也必须把JDBC捎上。当然反过来讲，不依赖数据的项目确实不多，貌似这么实现影响也不大。下面我们先根据官方文档配置一下看看效果。
首先修改POM配置，添加两个依赖。如果项目里本来已经使用了数据库，那这一步就不需要了。
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-jdbc</artifactId>
</dependency>
<dependency>
<groupId>com.h2database</groupId>
<artifactId>h2</artifactId>
<scope>runtime</scope>
</dependency>
然后修改RedisConfiguration
@Bean
public StringRedisTemplate redisTemplate(RedisConnectionFactory redisConnectionFactory) {
StringRedisTemplate template = new StringRedisTemplate();
template.setConnectionFactory(redisConnectionFactory);
template.setEnableTransactionSupport(true);//打开事务支持
return template;
}
//配置事务管理器
@Bean
public PlatformTransactionManager transactionManager(DataSource dataSource) throws SQLException {
return new DataSourceTransactionManager(dataSource);
}
我们新建一个RedisService，将原来的数据操作移到service里面。同时将Service方法加上@Transactional注解。
@Service
public class RedisService {
private StringRedisTemplate template;
public RedisService(StringRedisTemplate template) {
this.template = template;
}
@Transactional
public String put() {
int i = (int)(Math.random() * 100);
template.opsForValue().set("key"+i, "value"+i, 300, TimeUnit.SECONDS);
return "success "+"key"+i;
}
}
//-----------------------------------------------------------
//controller里面加一个新的方法，调用Service
@GetMapping("/puttx")
public String redisTxSet() {
return redisService.put();
}
完成这些工作之后，再往http://localhost:8080/puttx发送请求，无论点多少次，Redis的连接数始终维持在1个不变。在看程序的输出日志里面我们也发现了，事务结束后连接被正常释放。因为使用了JDBC的事务管理器，所以还顺便做了一次数据库事务的开启和提交。还有一点值得注意的是，跟数据库一样，使用注解来做事务管理，spring也会主动管理redis事务的提交和回滚，也就是在之前发送一条MULTI命令，成功后发送EXEC，失败后发送DISCARD。
2018-08-11 20:57:04.990 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [doGetConnection:126] Opening RedisConnection
2018-08-11 20:57:04.990 [DEBUG][http-nio-8080-exec-1]:o.springframework.aop.framework.JdkDynamicAopProxy [getProxy:118] Creating JDK dynamic proxy: target source is SingletonTargetSource for target object [org.springframework.data.redis.connection.jedis.JedisConnection@20f2be3c]
2018-08-11 20:57:04.990 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [intercept:337] Invoke 'multi' on bound conneciton
2018-08-11 20:57:04.991 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [intercept:337] Invoke 'isPipelined' on bound conneciton
2018-08-11 20:57:04.991 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [intercept:337] Invoke 'setEx' on bound conneciton
2018-08-11 20:57:04.991 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [releaseConnection:198] Redis Connection will be closed when transaction finished.
2018-08-11 20:57:04.991 [DEBUG][http-nio-8080-exec-1]:o.s.jdbc.datasource.DataSourceTransactionManager [processCommit:759] Initiating transaction commit
2018-08-11 20:57:04.991 [DEBUG][http-nio-8080-exec-1]:o.s.jdbc.datasource.DataSourceTransactionManager [doCommit:310] Committing JDBC transaction on Connection [ProxyConnection[PooledConnection[conn9: url=jdbc:h2:mem:testdb user=SA]]]
2018-08-11 20:57:04.992 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [intercept:337] Invoke 'exec' on bound conneciton
2018-08-11 20:57:04.992 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [afterCompletion:306] Closing bound connection after transaction completed with 0
2018-08-11 20:57:04.992 [DEBUG][http-nio-8080-exec-1]:o.s.data.redis.core.RedisConnectionUtils [intercept:337] Invoke 'close' on bound conneciton
2018-08-11 20:57:04.993 [DEBUG][http-nio-8080-exec-1]:o.s.jdbc.datasource.DataSourceTransactionManager [doCleanupAfterCompletion:368] Releasing JDBC Connection [ProxyConnection[PooledConnection[conn9: url=jdbc:h2:mem:testdb user=SA]]] after transaction
2018-08-11 20:57:04.993 [DEBUG][http-nio-8080-exec-1]:o.springframework.jdbc.datasource.DataSourceUtils [doReleaseConnection:327] Returning JDBC Connection to DataSource
总结
在spring中要使用Redis注解式事务，首先要设置RedisTemplate的enableTransactionSupport属性为true，然后配置一个jdbc的事务管理器。
这里有一点非常重要，一旦这样配置，所有使用这个template的redis操作都必须走注解式事务，要不然会导致连接一直占用，不关闭。
建议
- 升级到springboot 2.0以上版本，如果因为项目原因无法升级看下面的建议
- 如果使用Redis事务的场景不多，完全可以自己管理，不需要使用spring的注解式事务。如下面这样使用：
List<Object> txResults = redisTemplate.execute(new SessionCallback<List<Object>>() {
public List<Object> execute(RedisOperations operations) throws DataAccessException {
operations.multi();
operations.opsForSet().add("key", "value1");
// This will contain the results of all ops in the transaction
return operations.exec();
}
});
- 如果一定要使用spring提供的注解式事务，建议初始化两个
RedisTemplateBean，分别设置enableTransactionSupport属性为true和false。针对需要事务和不需要事务的操作使用不同的template。 - 从个人角度，我不建议使用redis事务，因为redis对于事务的支持并不是关系型数据库那样满足ACID。Redis事务只能保证ACID中的隔离性和一致性，无法保证原子性和持久性。而我们使用事务最重要的一个理由就是原子性，这一点无法保证，事务的意义就去掉一大半了。所以事务的场景可以尝试通过业务代码来实现。
