对应项目 springboot-matrix 下
seata-dubbo-interface
、seata-order-service
、seata-storage-service
、seata-account-service
在分布式系统中,一次操作由多个系统协同完成,这种一次操作协同多个系统完成的过程叫做分布式事务
在一个分布式系统中不可能同时满足一致性、可用性、分区容错性,最多满足两个,对于分布式系统而言,必须保证P,要么CP,要么AP
- 一致性 指强一致性,数据始终都是一致的
- 可用性 系统提供的服务一直处于可用状态,用户的操作请求在指定时间内响应请求,超出时间范围,认为系统不可用
- 分区容错性 分布式系统在遇到任何网络分区故障的时候,仍需要保证对外提供一致性和可用性,除非整个网络发生故障
网络分区 在分布系统中,可能由于故障导致节点之间无法连通,整个网络分成了几块区域
zookeeper采用的CP架构,eureka是AP架构,Nacos支持CP、AP架构
在cap理论上权衡的一种结果,即使做不到强一致性,但是可以做到最终一致性
- 基本可用
- 软状态
- 最终一致性
- 阶段一:提交事务请求(prepare)
1、协调者向所有参与者发送事务内容,询问是否可以提交,并等待参与者反馈
2、参与者执行本地事务,讲执行结果反馈给协调者
- 阶段二:执行事务提交(commit)
1、根据各参与者的反馈结果,是否执行事务提交,否则中断事务提交,回滚事务
2PC问题
-
同步阻塞
协调者要等待所有参与者反馈信息后才能继续执行
-
单点问题
协调者挂掉导致的事务一致性的问题
-
脑裂问题
网络分区问题,网络不好导致其中一个参与者挂掉事务不一致的问题
为了解决2PC的问题,引用了超时机制
-
阶段一 canCommit(准备阶段)
1、事务询问,是否可以执行事务
2、各参与者反馈事务询问的响应
-
阶段二 preCommit(预提交)
根据阶段一的反馈结果分为两种情况
1、 执行事务预提交
协调者向所有参与者发送preCommit请求,各参与者接收到预提交,执行事务操作,将执行结果反馈给协调者
2、中断事务
协调者向所有参与者发送abort请求,各参与者收到请求或者超时,都会中断事务
-
阶段三 doCommit(提交)
1、执行提交 协调者发送doCommit请求,各参与者收到请求执行提交事务,反馈ack给协调者,协调者收到各参与者ack,事务完成 2、中断事务 协调者向所有参与者发送abort请求,参与者收到请求,根据二阶段的undo信息进行事务回滚操作,回滚完成反馈ack消息,协调者收到 所有参与者ack消息,中断事务
3pc相较于2pc解决了阻塞和单点问题,但是还是无法解决脑裂问题
TCC补偿性事务是基于2PC实现的业务层事务控制方案,可以保证数据最终一致性
主要有三个方法,try、confirm、cancel
- 1、try检查及预留业务资源完成提交事务前检查,并预留资源
- 2、确认事务提交
- 3、取消事务提交,回滚事务
缺点
- 对代码侵入性强
- 需要在try、confirm接口实现幂等性(一个操作无论执行多少次,结果都是一样的)
基于消息队列+消息表实现的分布式事务,如rocketmq执行的分布式事务,保证数据的最终一致性
1、发送方给broker发送半消息,半消息对订阅方不可见,发送成功后执行本地事务,根据事务结果执行commit或者rollback, 发送方需要提交一个反查事务接口,用于broker检查事务是否执行成功
2、如果发送发commit成功,订阅方就可以收到这条消息,执行本地事务,消费消息
开源的分布式事务一致性解决方案,seata提供了AT、TCC、XA、SAGA事务模式
- TC (Transaction Coordinator) - 事务协调者:维护全局和分支事务的状态,驱动全局事务提交或回滚。
- TM (Transaction Manager) - 事务管理器:定义全局事务的范围,开始全局事务、提交或回滚全局事务。
- RM ( Resource Manager ) - 资源管理器:管理分支事务处理的资源( Resource ),与 TC 交谈以注册分支事务和报告分支事务的状态,并驱动分支事务提交或回滚。
AT 模式是一种无侵入的分布式事务解决方案。在 AT 模式下,用户只需关注自己的“业务 SQL”,用户的 “业务 SQL” 作为一阶段,Seata 框架会自动生成事务的二阶段提交和回滚操作。
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。
一阶段
seata会拦截sql,解析元数据,在执行sql前保存before image、然后执行sql,执行sql后保存after image,讲前后镜像数据以及业务sql组成一条回滚日志记录,插入到undo表中, 生成行锁,以上全部在一个事务中完成,保证一阶段的原子性
二阶段提交
如是二阶段是提交的话,因为sql已经在一阶段提交到数据库了,所以只需要删除一阶段保存的快照和undo log即可,请求会放在异步队列中执行
二阶段回滚
如果二阶段是回滚,则根据一阶段保存的快照before image还原业务数据,在还原之前,需要根据对比当前数据和after image中的数据是否一致, 如果一致则没有脏写,可以还原,或者根据配置策略处理
写隔离
- 阶段一在事务提交前,需要拿到全局锁
- 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。
读隔离
- seata默认隔离级别是读未提交
- 如需要求读已提交,使用select FROM UPDATE语句
使用版本,springboot2.5.6,seata-spring-boot-starter 1.4.2,dubbo-spring-boot-starter3.0.4,mysql:mysql-connector-java 8.0.22
nacos服务器版本:2.0.2, seata服务器版本:1.4.2
-
下载nacos,以单机模式部署nacos
-
下载seata,更新conf目录下的file.conf和registry.conf配置
我的配置如下
file.cnf
## transaction log store, only used in seata-server
store {
## store mode: file、db、redis
mode = "db"
## rsa decryption public key
publicKey = ""
## file store property
file {
## store location dir
dir = "sessionStore"
# branch session size, if exceeded first try compress lockkey, still exceeded throws exceptions
maxBranchSessionSize = 16384
# globe session size, if exceeded throws exceptions
maxGlobalSessionSize = 512
# file buffer size, if exceeded allocate new buffer
fileWriteBufferCacheSize = 16384
# when recover batch read size
sessionReloadReadSize = 100
# async, sync
flushDiskMode = async
}
## database store property
db {
## the implement of javax.sql.DataSource, such as DruidDataSource(druid)/BasicDataSource(dbcp)/HikariDataSource(hikari) etc.
datasource = "druid"
## mysql/oracle/postgresql/h2/oceanbase etc.
dbType = "mysql"
driverClassName = "com.mysql.cj.jdbc.Driver"
## if using mysql to store the data, recommend add rewriteBatchedStatements=true in jdbc connection param
url = "jdbc:mysql://127.0.0.1:3306/seata?rewriteBatchedStatements=true"
user = "root"
password = "123456"
minConn = 5
maxConn = 100
globalTable = "global_table"
branchTable = "branch_table"
lockTable = "lock_table"
queryLimit = 100
maxWait = 5000
}
## redis store property
redis {
## redis mode: single、sentinel
mode = "single"
## single mode property
single {
host = "127.0.0.1"
port = "6379"
}
## sentinel mode property
sentinel {
masterName = ""
## such as "10.28.235.65:26379,10.28.235.65:26380,10.28.235.65:26381"
sentinelHosts = ""
}
password = ""
database = "0"
minConn = 1
maxConn = 10
maxTotal = 100
queryLimit = 100
}
}
registry.conf
registry {
# file 、nacos 、eureka、redis、zk、consul、etcd3、sofa
type = "nacos"
nacos {
application = "seata-server"
serverAddr = "127.0.0.1:8848"
group = "SEATA_GROUP"
namespace = "seata"
cluster = "default"
username = "nacos"
password = "nacos"
}
eureka {
serviceUrl = "http://localhost:8761/eureka"
application = "default"
weight = "1"
}
redis {
serverAddr = "localhost:6379"
db = 0
password = ""
cluster = "default"
timeout = 0
}
zk {
cluster = "default"
serverAddr = "127.0.0.1:2181"
sessionTimeout = 6000
connectTimeout = 2000
username = ""
password = ""
}
consul {
cluster = "default"
serverAddr = "127.0.0.1:8500"
aclToken = ""
}
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 = "nacos"
nacos {
serverAddr = "127.0.0.1:8848"
namespace = "seata"
group = "SEATA_GROUP"
username = "nacos"
password = "nacos"
dataId = "seataServer.properties"
}
consul {
serverAddr = "127.0.0.1:8500"
aclToken = ""
}
apollo {
appId = "seata-server"
## apolloConfigService will cover apolloMeta
apolloMeta = "http://192.168.1.204:8801"
apolloConfigService = "http://192.168.1.204:8080"
namespace = "application"
apolloAccesskeySecret = ""
cluster = "seata"
}
zk {
serverAddr = "127.0.0.1:2181"
sessionTimeout = 6000
connectTimeout = 2000
username = ""
password = ""
nodePath = "/seata/seata.properties"
}
etcd3 {
serverAddr = "http://localhost:2379"
}
file {
name = "file.conf"
}
}
创建seata
数据库,执行sql脚本
-- -------------------------------- The script used when storeMode is 'db' --------------------------------
-- the table to store GlobalSession data
CREATE TABLE IF NOT EXISTS `global_table`
(
`xid` VARCHAR(128) NOT NULL,
`transaction_id` BIGINT,
`status` TINYINT NOT NULL,
`application_id` VARCHAR(32),
`transaction_service_group` VARCHAR(32),
`transaction_name` VARCHAR(128),
`timeout` INT,
`begin_time` BIGINT,
`application_data` VARCHAR(2000),
`gmt_create` DATETIME,
`gmt_modified` DATETIME,
PRIMARY KEY (`xid`),
KEY `idx_gmt_modified_status` (`gmt_modified`, `status`),
KEY `idx_transaction_id` (`transaction_id`)
) ENGINE = InnoDB
DEFAULT CHARSET = utf8;
-- the table to store BranchSession data
CREATE TABLE IF NOT EXISTS `branch_table`
(
`branch_id` BIGINT NOT NULL,
`xid` VARCHAR(128) NOT NULL,
`transaction_id` BIGINT,
`resource_group_id` VARCHAR(32),
`resource_id` VARCHAR(256),
`branch_type` VARCHAR(8),
`status` TINYINT,
`client_id` VARCHAR(64),
`application_data` VARCHAR(2000),
`gmt_create` DATETIME(6),
`gmt_modified` DATETIME(6),
PRIMARY KEY (`branch_id`),
KEY `idx_xid` (`xid`)
) ENGINE = InnoDB
DEFAULT CHARSET = utf8;
-- the table to store lock data
CREATE TABLE IF NOT EXISTS `lock_table`
(
`row_key` VARCHAR(128) NOT NULL,
`xid` VARCHAR(128),
`transaction_id` BIGINT,
`branch_id` BIGINT NOT NULL,
`resource_id` VARCHAR(256),
`table_name` VARCHAR(32),
`pk` VARCHAR(36),
`gmt_create` DATETIME,
`gmt_modified` DATETIME,
PRIMARY KEY (`row_key`),
KEY `idx_branch_id` (`branch_id`)
) ENGINE = InnoDB
DEFAULT CHARSET = utf8;
CREATE TABLE IF NOT EXISTS `distributed_lock`
(
`lock_key` CHAR(20) NOT NULL,
`lock_value` VARCHAR(20) NOT NULL,
`expire` BIGINT,
primary key (`lock_key`)
) ENGINE = InnoDB
DEFAULT CHARSET = utf8mb4;
INSERT INTO `distributed_lock` (lock_key, lock_value, expire)
VALUES ('AsyncCommitting', ' ', 0);
INSERT INTO `distributed_lock` (lock_key, lock_value, expire)
VALUES ('RetryCommitting', ' ', 0);
INSERT INTO `distributed_lock` (lock_key, lock_value, expire)
VALUES ('RetryRollbacking', ' ', 0);
INSERT INTO `distributed_lock` (lock_key, lock_value, expire)
VALUES ('TxTimeoutCheck', ' ', 0);
在各业务系统database上创建undo_log
表
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;
- 启动nacos、seata服务
- 创建三个服务,order-service、storage-service、account-service服务模块
- gradle依赖如下
dependencies {
implementation(project(":seata-dubbo-interface"))
implementation("com.baomidou:mybatis-plus-boot-starter")
implementation 'org.springframework.boot:spring-boot-starter-web'
implementation("mysql:mysql-connector-java")
implementation 'org.apache.dubbo:dubbo-spring-boot-starter'
implementation 'org.apache.dubbo:dubbo-registry-nacos'
implementation 'io.seata:seata-spring-boot-starter'
testImplementation("org.springframework.boot:spring-boot-starter-test")
}
application.yml配置文件如下,这里只展示order服务的配置
spring:
application:
name: order-service
datasource:
url: jdbc:mysql://localhost:3306/seata_order?allowPublicKeyRetrieval=true&useSSL=false&useUnicode=true&characterEncoding=UTF-8&autoReconnect=true&useSSL=false&zeroDateTimeBehavior=convertToNull&serverTimezone=Asia/Shanghai
driver-class-name: com.mysql.cj.jdbc.Driver
username: root
password: 123456
server:
port: 8090
mybatis-plus:
mapper-locations: classpath*:/mapper/*.xml
typeAliasesPackage: com.zwl.domain
configuration:
map-underscore-to-camel-case: true
global-config:
db-config:
id-type: auto
#dubbo配置
dubbo:
application:
id: ${spring.application.name}
name: ${spring.application.name}
qos-enable: false
version: 1.0.0
registry:
id: ${spring.application.name}-registry
address: nacos://127.0.0.1:8848
username: nacos
password: nacos
check: true
simplified: true
config-center:
address: nacos://127.0.0.1:8848
check: true
username: nacos
password: nacos
metadata-report:
address: nacos://127.0.0.1:8848
username: nacos
password: nacos
protocol:
id: dubbo
name: dubbo
port: -1
provider:
timeout: 1000
version: 1.0.0
consumer:
timeout: 1000
version: 1.0.0
scan:
base-packages: com.zwl
#seata配置
seata:
enabled: true
data-source-proxy-mode: AT
enable-auto-data-source-proxy: true
application-id: ${spring.application.name}
tx-service-group: ${spring.application.name}-group
service:
vgroup-mapping:
order-service-group: default
grouplist:
default: 127.0.0.1:8091
registry:
type: nacos
nacos:
cluster: default
application: seata-server
server-addr: 127.0.0.1:8848
namespace: seata
group: SEATA_GROUP
username: nacos
password: nacos