Seata（simple extensiable autonomous transaction architecture）

Seata 是一款开源的分布式事务解决方案，致力于提供高性能和简单易用的分布式事务服务。Seata 将为用户提供了 AT、TCC、SAGA 和 XA 事务模式，为用户打造一站式的分布式解决方案。

AT 模式（比较完善）

前提

基于支持本地 ACID 事务的关系型数据库。
Java 应用，通过 JDBC 访问数据库。

整体机制

两阶段提交协议的演变：

一阶段：业务数据和回滚日志记录在同一个本地事务中提交，释放本地锁和连接资源。
二阶段：
- 提交异步化，非常快速地完成。
- 回滚通过一阶段的回滚日志进行反向补偿。

写隔离

一阶段本地事务提交前，需要确保先拿到 全局锁 。
拿不到 全局锁 ，不能提交本地事务。
拿 全局锁 的尝试被限制在一定范围内，超出范围将放弃，并回滚本地事务，释放本地锁。

以一个示例来说明：

两个全局事务 tx1 和 tx2，分别对 a 表的 m 字段进行更新操作，m 的初始值 1000。

tx1 先开始，开启本地事务，拿到本地锁，更新操作 m = 1000 - 100 = 900。本地事务提交前，先拿到该记录的 全局锁 ，本地提交释放本地锁。 tx2 后开始，开启本地事务，拿到本地锁，更新操作 m = 900 - 100 = 800。本地事务提交前，尝试拿该记录的 全局锁 ，tx1 全局提交前，该记录的全局锁被 tx1 持有，tx2 需要重试等待 全局锁 。

如果 tx1 的二阶段全局回滚，则 tx1 需要重新获取该数据的本地锁，进行反向补偿的更新操作，实现分支的回滚。

此时，如果 tx2 仍在等待该数据的 全局锁，同时持有本地锁，则 tx1 的分支回滚会失败。分支的回滚会一直重试，直到 tx2 的 全局锁 等锁超时，放弃 全局锁 并回滚本地事务释放本地锁，tx1 的分支回滚最终成功。

因为整个过程 全局锁 在 tx1 结束前一直是被 tx1 持有的，所以不会发生脏写的问题。

读隔离

在数据库本地事务隔离级别 读已提交（Read Committed） 或以上的基础上，Seata（AT 模式）的默认全局隔离级别是 读未提交（Read Uncommitted） 。

如果应用在特定场景下，必需要求全局的 读已提交 ，目前 Seata 的方式是通过 SELECT FOR UPDATE 语句的代理。

SELECT FOR UPDATE 语句的执行会申请 全局锁 ，如果 全局锁 被其他事务持有，则释放本地锁（回滚 SELECT FOR UPDATE 语句的本地执行）并重试。这个过程中，查询是被 block 住的，直到 全局锁 拿到，即读取的相关数据是 已提交 的，才返回。

出于总体性能上的考虑，Seata 目前的方案并没有对所有 SELECT 语句都进行代理，仅针对 FOR UPDATE 的 SELECT 语句。

工作机制

以一个示例来说明整个 AT 分支的工作过程。

业务表：product

AT 分支事务的业务逻辑：

update product set name = 'GTS' where name = 'TXC';

一阶段

过程：

解析 SQL：得到 SQL 的类型（UPDATE），表（product），条件（where name = 'TXC'）等相关的信息。
查询前镜像：根据解析得到的条件信息，生成查询语句，定位数据

 select id, name, since from product where name = 'TXC';


得到前镜像：

执行业务 SQL：更新这条记录的 name 为 'GTS'。
查询后镜像：根据前镜像的结果，通过主键定位数据。

                select id, name, since from product where id = 1;
得到后镜像：

5.插入回滚日志：把前后镜像数据以及业务 SQL 相关的信息组成一条回滚日志记录，插入到 UNDO_LOG 表中。

{
    "branchId": 641789253,
    "undoItems": [{
        "afterImage": {
            "rows": [{
                "fields": [{
                    "name": "id",
                    "type": 4,
                    "value": 1
                }, {
                    "name": "name",
                    "type": 12,
                    "value": "GTS"
                }, {
                    "name": "since",
                    "type": 12,
                    "value": "2014"
                }]
            }],
            "tableName": "product"
        },
        "beforeImage": {
            "rows": [{
                "fields": [{
                    "name": "id",
                    "type": 4,
                    "value": 1
                }, {
                    "name": "name",
                    "type": 12,
                    "value": "TXC"
                }, {
                    "name": "since",
                    "type": 12,
                    "value": "2014"
                }]
            }],
            "tableName": "product"
        },
        "sqlType": "UPDATE"
    }],
    "xid": "xid:xxx"
}

提交前，向 TC 注册分支：申请 product 表中，主键值等于 1 的记录的 全局锁 。
本地事务提交：业务数据的更新和前面步骤中生成的 UNDO LOG 一并提交。
将本地事务提交的结果上报给 TC。

二阶段-回滚

收到 TC 的分支回滚请求，开启一个本地事务，执行如下操作。
通过 XID 和 Branch ID 查找到相应的 UNDO LOG 记录。
数据校验：拿 UNDO LOG 中的后镜与当前数据进行比较，如果有不同，说明数据被当前全局事务之外的动作做了修改。这种情况，需要根据配置策略来做处理，详细的说明在另外的文档中介绍。
根据 UNDO LOG 中的前镜像和业务 SQL 的相关信息生成并执行回滚的语句：

update product set name = 'TXC' where id = 1;

提交本地事务。并把本地事务的执行结果（即分支事务回滚的结果）上报给 TC。

二阶段-提交

收到 TC 的分支提交请求，把请求放入一个异步任务的队列中，马上返回提交成功的结果给 TC。
异步任务阶段的分支提交请求将异步和批量地删除相应 UNDO LOG 记录。

-- 注意此处0.7.0+ 增加字段 context
CREATE TABLE `undo_log` (
  `id` bigint(20) NOT NULL AUTO_INCREMENT,
  `branch_id` bigint(20) NOT NULL,
  `xid` varchar(100) NOT NULL,
  `context` varchar(128) NOT NULL,
  `rollback_info` longblob NOT NULL,
  `log_status` int(11) NOT NULL,
  `log_created` datetime NOT NULL,
  `log_modified` datetime NOT NULL,
  PRIMARY KEY (`id`),
  UNIQUE KEY `ux_undo_log` (`xid`,`branch_id`)
) ENGINE=InnoDB AUTO_INCREMENT=1 DEFAULT CHARSET=utf8;

------------------------------------------------------------------------实现------------------------------------------------------

1.下载seata-server-0.9.0并解压

2.修改registry.conf文件

registry {
  # file 、nacos 、eureka、redis、zk、consul、etcd3、sofa
  type = "file"   

  nacos {
    serverAddr = "localhost"
    namespace = ""
    cluster = "default"
  }
  eureka {
    serviceUrl = "http://localhost:8761/eureka"
    application = "default"
    weight = "1"
  }
  redis {
    serverAddr = "localhost:6379"
    db = "0"
  }
  zk {
    cluster = "default"
    serverAddr = "127.0.0.1:2181"
    session.timeout = 6000
    connect.timeout = 2000
  }
  consul {
    cluster = "default"
    serverAddr = "127.0.0.1:8500"
  }
  etcd3 {
    cluster = "default"
    serverAddr = "http://localhost:2379"
  }
  sofa {
    serverAddr = "127.0.0.1:9603"
    application = "default"
    region = "DEFAULT_ZONE"
    datacenter = "DefaultDataCenter"
    cluster = "default"
    group = "SEATA_GROUP"
    addressWaitTime = "3000"
  }
  file {
    name = "file.conf"
  }
}

config {
  # file、nacos 、apollo、zk、consul、etcd3
  type = "file"

  nacos {
    serverAddr = "localhost"
    namespace = ""
  }
  consul {
    serverAddr = "127.0.0.1:8500"
  }
  apollo {
    app.id = "seata-server"
    apollo.meta = "http://192.168.1.204:8801"
  }
  zk {
    serverAddr = "127.0.0.1:2181"
    session.timeout = 6000
    connect.timeout = 2000
  }
  etcd3 {
    serverAddr = "http://localhost:2379"
  }
  file {
    name = "file.conf"
  }
}

3.有必要的时候修改file.conf 目前没有发挥作用

transport {
  # tcp udt unix-domain-socket
  type = "TCP"
  #NIO NATIVE
  server = "NIO"
  #enable heartbeat
  heartbeat = true
  #thread factory for netty
  thread-factory {
    boss-thread-prefix = "NettyBoss"
    worker-thread-prefix = "NettyServerNIOWorker"
    server-executor-thread-prefix = "NettyServerBizHandler"
    share-boss-worker = false
    client-selector-thread-prefix = "NettyClientSelector"
    client-selector-thread-size = 1
    client-worker-thread-prefix = "NettyClientWorkerThread"
    # netty boss thread size,will not be used for UDT
    boss-thread-size = 1
    #auto default pin or 8
    worker-thread-size = 8
  }
  shutdown {
    # when destroy server, wait seconds
    wait = 3
  }
  serialization = "seata"
  compressor = "none"
}
service {
  #vgroup->rgroup
  vgroup_mapping.my_test_tx_group = "default"
  #only support single node
  default.grouplist = "127.0.0.1:8091"
  #degrade current not support
  enableDegrade = false
  #disable
  disable = false
  #unit ms,s,m,h,d represents milliseconds, seconds, minutes, hours, days, default permanent
  max.commit.retry.timeout = "-1"
  max.rollback.retry.timeout = "-1"
}

client {
  async.commit.buffer.limit = 10000
  lock {
    retry.internal = 10
    retry.times = 30
  }
  report.retry.count = 5
  tm.commit.retry.count = 1
  tm.rollback.retry.count = 1
}

## transaction log store
store {
  ## store mode: file、db
  mode = "file"

  ## file store
  file {
    dir = "sessionStore"

    # branch session size , if exceeded first try compress lockkey, still exceeded throws exceptions
    max-branch-session-size = 16384
    # globe session size , if exceeded throws exceptions
    max-global-session-size = 512
    # file buffer size , if exceeded allocate new buffer
    file-write-buffer-cache-size = 16384
    # when recover batch read size
    session.reload.read_size = 100
    # async, sync
    flush-disk-mode = async
  }

  ## database store
  db {
    ## the implement of javax.sql.DataSource, such as DruidDataSource(druid)/BasicDataSource(dbcp) etc.
    datasource = "dbcp"
    ## mysql/oracle/h2/oceanbase etc.
    db-type = "mysql"
    driver-class-name = "com.mysql.jdbc.Driver"
    url = "jdbc:mysql://127.0.0.1:3306/seata"
    user = "mysql"
    password = "mysql"
    min-conn = 1
    max-conn = 3
##这几个表需要在数据库新建，文件在conf文件夹下
    global.table = "global_table"
    branch.table = "branch_table"
    lock-table = "lock_table"
    query-limit = 100
  }
}
lock {
  ## the lock store mode: local、remote
  mode = "remote"

  local {
    ## store locks in user's database
  }

  remote {
    ## store locks in the seata's server
  }
}
recovery {
  #schedule committing retry period in milliseconds
  committing-retry-period = 1000
  #schedule asyn committing retry period in milliseconds
  asyn-committing-retry-period = 1000
  #schedule rollbacking retry period in milliseconds
  rollbacking-retry-period = 1000
  #schedule timeout retry period in milliseconds
  timeout-retry-period = 1000
}

transaction {
  undo.data.validation = true
  undo.log.serialization = "jackson"
  undo.log.save.days = 7
  #schedule delete expired undo_log in milliseconds
  undo.log.delete.period = 86400000
  undo.log.table = "undo_log"  ##必须存在，那个微服务需要用到seata就在那个微服务所在库新建，文件也在conf文件夹下
}

## metrics settings
metrics {
  enabled = false
  registry-type = "compact"
  # multi exporters use comma divided
  exporter-list = "prometheus"
  exporter-prometheus-port = 9898
}

support {
  ## spring
  spring {
    # auto proxy the DataSource bean
    datasource.autoproxy = false
  }
}

4.可以去bin目录下启动seata

5.在微服务中新建那个微服务需要用到seata就在那个微服务下创建最好抽取出来

package com.atguigu.gmall.pms.config;

import com.zaxxer.hikari.HikariDataSource;
import io.seata.rm.datasource.DataSourceProxy;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.context.annotation.Primary;
import javax.sql.DataSource;

@Configuration
public class DataSourceConfig {

    /**
     * 需要将 DataSourceProxy 设置为主数据源，否则事务无法回滚
     */
    @Primary
    @Bean("dataSource")
    public DataSource dataSource(@Value("${spring.datasource.url}")String url,@Value("${spring.datasource.driver-class-name}")String driverClassName,
                                 @Value("${spring.datasource.username}")String username,@Value("${spring.datasource.password}")String password) {
        HikariDataSource hikariDataSource = new HikariDataSource();
        hikariDataSource.setJdbcUrl(url);
        hikariDataSource.setDriverClassName(driverClassName);
        hikariDataSource.setUsername(username);
        hikariDataSource.setPassword(password);
        return new DataSourceProxy(hikariDataSource);
    }
}

在resource下新建file.conf registry.conf 文件，，，也在解压后seata文件中有 conf目录下

registry.conf文件

相应选择:

registry {
  # file 、nacos 、eureka、redis、zk
  type = "nacos"

  nacos {
    serverAddr = "localhost:8848"
    namespace = "public"
    cluster = "default"
  }

}

config {
  # file、nacos 、apollo、zk
  type = "file"

  file {
    name = "file.conf"
  }
}

file.conf文件

transport {
  # tcp udt unix-domain-socket
  type = "TCP"
  #NIO NATIVE
  server = "NIO"
  #enable heartbeat
  heartbeat = true
  #thread factory for netty
  thread-factory {
    boss-thread-prefix = "NettyBoss"
    worker-thread-prefix = "NettyServerNIOWorker"
    server-executor-thread-prefix = "NettyServerBizHandler"
    share-boss-worker = false
    client-selector-thread-prefix = "NettyClientSelector"
    client-selector-thread-size = 1
    client-worker-thread-prefix = "NettyClientWorkerThread"
    # netty boss thread size,will not be used for UDT
    boss-thread-size = 1
    #auto default pin or 8
    worker-thread-size = 8
  }
  shutdown {
    # when destroy server, wait seconds
    wait = 3
  }
  serialization = "seata"
  compressor = "none"
}
service {
  #vgroup->rgroup
#默认是vgroup_mapping.${spring.application.name}-fescar-service-group   ,,  vgroup_mapping.这部分可以在配置文件中配置
  vgroup_mapping.pms-server-fescar-service-group="default"  
  #only support single node
  default.grouplist = "127.0.0.1:8091"
  #degrade current not support
  enableDegrade = false
  #disable
  disable = false
  #unit ms,s,m,h,d represents milliseconds, seconds, minutes, hours, days, default permanent
  max.commit.retry.timeout = "-1"
  max.rollback.retry.timeout = "-1"
}

client {
  async.commit.buffer.limit = 10000
  lock {
    retry.internal = 10
    retry.times = 30
  }
  report.retry.count = 5
  tm.commit.retry.count = 1
  tm.rollback.retry.count = 1
}

## transaction log store
store {
  ## store mode: file、db
  mode = "file"   #推荐选择file，，db好像还不完善，，，下面这些链接随便写也不会报错，目前

  ## file store
  file {
    dir = "sessionStore"

    # branch session size , if exceeded first try compress lockkey, still exceeded throws exceptions
    max-branch-session-size = 16384
    # globe session size , if exceeded throws exceptions
    max-global-session-size = 512
    # file buffer size , if exceeded allocate new buffer
    file-write-buffer-cache-size = 16384
    # when recover batch read size
    session.reload.read_size = 100
    # async, sync
    flush-disk-mode = async
  }

  ## database store
  db {
    ## the implement of javax.sql.DataSource, such as DruidDataSource(druid)/BasicDataSource(dbcp) etc.
    datasource = "dbcp"
    ## mysql/oracle/h2/oceanbase etc.
    db-type = "mysql"
    driver-class-name = "com.mysql.cj.jdbc.Driver"
    url = "jdbc:mysql://192.168.146.140:3306/seata"
    user = "root"
    password = "123456"
    min-conn = 1
    max-conn = 3
    global.table = "global_table"
    branch.table = "branch_table"
    lock-table = "lock_table"
    query-limit = 100
  }
}
lock {
  ## the lock store mode: local、remote
  mode = "remote"

  local {
    ## store locks in user's database
  }

  remote {
    ## store locks in the seata's server
  }
}
recovery {
  #schedule committing retry period in milliseconds
  committing-retry-period = 1000
  #schedule asyn committing retry period in milliseconds
  asyn-committing-retry-period = 1000
  #schedule rollbacking retry period in milliseconds
  rollbacking-retry-period = 1000
  #schedule timeout retry period in milliseconds
  timeout-retry-period = 1000
}

transaction {
  undo.data.validation = true
  undo.log.serialization = "jackson"
  undo.log.save.days = 7
  #schedule delete expired undo_log in milliseconds
  undo.log.delete.period = 86400000
  undo.log.table = "undo_log"
}

## metrics settings
metrics {
  enabled = false
  registry-type = "compact"
  # multi exporters use comma divided
  exporter-list = "prometheus"
  exporter-prometheus-port = 9898
}

support {
  ## spring
  spring {
    # auto proxy the DataSource bean
    datasource.autoproxy = false
  }
}

6.在主入口中添加

    @Override
    @GlobalTransactional
    public void master(...) {}


选项：
// Source code recreated from a .class file by IntelliJ IDEA
// (powered by Fernflower decompiler)
//

package io.seata.spring.annotation;

import java.lang.annotation.ElementType;
import java.lang.annotation.Inherited;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;

@Retention(RetentionPolicy.RUNTIME)
@Target({ElementType.METHOD})
@Inherited
public @interface GlobalTransactional {
    int timeoutMills() default 60000;

    String name() default "";

    Class<? extends Throwable>[] rollbackFor() default {};

    String[] rollbackForClassName() default {};

    Class<? extends Throwable>[] noRollbackFor() default {};

    String[] noRollbackForClassName() default {};
}

其他进口：

    @Override
    @Transactional
    public void saveSkuSaleInfo(SkuSlaseVo skuSlaseVo) {}


选项:
// Source code recreated from a .class file by IntelliJ IDEA
// (powered by Fernflower decompiler)
//

package org.springframework.transaction.annotation;

import java.lang.annotation.Documented;
import java.lang.annotation.ElementType;
import java.lang.annotation.Inherited;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;
import org.springframework.core.annotation.AliasFor;

@Target({ElementType.TYPE, ElementType.METHOD})
@Retention(RetentionPolicy.RUNTIME)
@Inherited
@Documented
public @interface Transactional {
    @AliasFor("transactionManager")
    String value() default "";

    @AliasFor("value")
    String transactionManager() default "";

    Propagation propagation() default Propagation.REQUIRED;

    Isolation isolation() default Isolation.DEFAULT;

    int timeout() default -1;

    boolean readOnly() default false;

    Class<? extends Throwable>[] rollbackFor() default {};

    String[] rollbackForClassName() default {};

    Class<? extends Throwable>[] noRollbackFor() default {};

    String[] noRollbackForClassName() default {};
}

报错：Caused by: io.seata.common.exception.ShouldNeverHappenException: Could not found any index in the table:**** seata默认每张表都要有主键

官网：http://seata.io

posted on 2020-07-14 19:03 lvym777 阅读(490) 评论(1) 编辑收藏举报

刷新页面返回顶部