前言： 坑点太多了，以至于需要单独写篇博客记录一下。
网上教程五花八门且不声明版本，文档不对应以及seata本身的bug，就造成了部署时各种踩坑，如果你和博主一样，已经又恰好很久没碰过nacos了，那可能对些配置就更陌生了，为尽可能避免部署，项目启动、运行后报错 或者服务重启后开始报错，请尽可能按照博主步骤操作，下面进入正题。
注：本文暂未博包含高可用教程，适用于demo及开发环境。
本文由csdn博主：孟秋与你 编写 ,花费许多精力排坑，主打一个让小白也能轻松入门，转载请声明出处
文章目录
环境
本文环境采用的是2021.x中最新的版本，也是springboot 2中 最新的版本对应关系,
亲测可行,阅读教程前请先核对自己的版本。
（2022.x与springboot3对应，而springboot3又和高版本jdk关联，博主暂没研究）
nacos版本：2.2.0
seata版本：1.6.1
springboot: 2.6.13
spring-cloud: 2021.0.5
spring-cloud-alibaba: 2021.0.5.0
其它：
mybatis plus: 3.5.1
openfeign: 3.1.5
loadbalancer: 3.1.5 (openfeign要用到)
nacos服务端配置
- 修改nacos/application.properties文件
主要就是改一下数据库的东西，比较简单
#*************** Spring Boot Related Configurations ***************#
### Default web context path:
server.servlet.contextPath=/nacos
### Default web server port:
server.port=8848
#*************** Network Related Configurations ***************#
### If prefer hostname over ip for Nacos server addresses in cluster.conf:
# nacos.inetutils.prefer-hostname-over-ip=false
### Specify local server's IP:
# nacos.inetutils.ip-address=
#*************** Config Module Related Configurations ***************#
### If use MySQL as datasource:
# spring.datasource.platform=mysql
### Count of DB:
db.num=1
### Connect URL of DB:
db.url.0=jdbc:mysql://127.0.0.1:3306/test?characterEncoding=utf8&connectTimeout=1000&socketTimeout=3000&autoReconnect=true&useUnicode=true&useSSL=false&serverTimezone=UTC
db.user.0=root
db.password.0=root
### Connection pool configuration: hikariCP
db.pool.config.connectionTimeout=30000
db.pool.config.validationTimeout=10000
db.pool.config.maximumPoolSize=20
db.pool.config.minimumIdle=2
#*************** Naming Module Related Configurations ***************#
### Data dispatch task execution period in milliseconds: Will removed on v2.1.X, replace with nacos.core.protocol.distro.data.sync.delayMs
# nacos.naming.distro.taskDispatchPeriod=200
### Data count of batch sync task: Will removed on v2.1.X. Deprecated
# nacos.naming.distro.batchSyncKeyCount=1000
### Retry delay in milliseconds if sync task failed: Will removed on v2.1.X, replace with nacos.core.protocol.distro.data.sync.retryDelayMs
# nacos.naming.distro.syncRetryDelay=5000
### If enable data warmup. If set to false, the server would accept request without local data preparation:
# nacos.naming.data.warmup=true
### If enable the instance auto expiration, kind like of health check of instance:
# nacos.naming.expireInstance=true
### will be removed and replaced by `nacos.naming.clean` properties
nacos.naming.empty-service.auto-clean=true
nacos.naming.empty-service.clean.initial-delay-ms=50000
nacos.naming.empty-service.clean.period-time-ms=30000
### Add in 2.0.0
### The interval to clean empty service, unit: milliseconds.
# nacos.naming.clean.empty-service.interval=60000
### The expired time to clean empty service, unit: milliseconds.
# nacos.naming.clean.empty-service.expired-time=60000
### The interval to clean expired metadata, unit: milliseconds.
# nacos.naming.clean.expired-metadata.interval=5000
### The expired time to clean metadata, unit: milliseconds.
# nacos.naming.clean.expired-metadata.expired-time=60000
### The delay time before push task to execute from service changed, unit: milliseconds.
# nacos.naming.push.pushTaskDelay=500
### The timeout for push task execute, unit: milliseconds.
# nacos.naming.push.pushTaskTimeout=5000
### The delay time for retrying failed push task, unit: milliseconds.
# nacos.naming.push.pushTaskRetryDelay=1000
### Since 2.0.3
### The expired time for inactive client, unit: milliseconds.
# nacos.naming.client.expired.time=180000
#*************** CMDB Module Related Configurations ***************#
### The interval to dump external CMDB in seconds:
# nacos.cmdb.dumpTaskInterval=3600
### The interval of polling data change event in seconds:
# nacos.cmdb.eventTaskInterval=10
### The interval of loading labels in seconds:
# nacos.cmdb.labelTaskInterval=300
### If turn on data loading task:
# nacos.cmdb.loadDataAtStart=false
#*************** Metrics Related Configurations ***************#
### Metrics for prometheus
#management.endpoints.web.exposure.include=*
### Metrics for elastic search
management.metrics.export.elastic.enabled=false
#management.metrics.export.elastic.host=http://localhost:9200
### Metrics for influx
management.metrics.export.influx.enabled=false
#management.metrics.export.influx.db=springboot
#management.metrics.export.influx.uri=http://localhost:8086
#management.metrics.export.influx.auto-create-db=true
#management.metrics.export.influx.consistency=one
#management.metrics.export.influx.compressed=true
#*************** Access Log Related Configurations ***************#
### If turn on the access log:
server.tomcat.accesslog.enabled=true
### The access log pattern:
server.tomcat.accesslog.pattern=%h %l %u %t "%r" %s %b %D %{User-Agent}i %{Request-Source}i
### The directory of access log:
server.tomcat.basedir=
#*************** Access Control Related Configurations ***************#
### If enable spring security, this option is deprecated in 1.2.0:
#spring.security.enabled=false
### The ignore urls of auth, is deprecated in 1.2.0:
nacos.security.ignore.urls=/,/error,/**/*.css,/**/*.js,/**/*.html,/**/*.map,/**/*.svg,/**/*.png,/**/*.ico,/console-ui/public/**,/v1/auth/**,/v1/console/health/**,/actuator/**,/v1/console/server/**
### The auth system to use, currently only 'nacos' and 'ldap' is supported:
nacos.core.auth.system.type=nacos
### If turn on auth system:
nacos.core.auth.enabled=false
### worked when nacos.core.auth.system.type=ldap，{0} is Placeholder,replace login username
# nacos.core.auth.ldap.url=ldap://localhost:389
# nacos.core.auth.ldap.userdn=cn={0},ou=user,dc=company,dc=com
### The token expiration in seconds:
nacos.core.auth.default.token.expire.seconds=18000
### The default token:
nacos.core.auth.default.token.secret.key=SecretKey012345678901234567890123456789012345678901234567890123456789
### Turn on/off caching of auth information. By turning on this switch, the update of auth information would have a 15 seconds delay.
nacos.core.auth.caching.enabled=true
### Since 1.4.1, Turn on/off white auth for user-agent: nacos-server, only for upgrade from old version.
nacos.core.auth.enable.userAgentAuthWhite=false
### Since 1.4.1, worked when nacos.core.auth.enabled=true and nacos.core.auth.enable.userAgentAuthWhite=false.
### The two properties is the white list for auth and used by identity the request from other server.
nacos.core.auth.server.identity.key=serverIdentity
nacos.core.auth.server.identity.value=security
#*************** Istio Related Configurations ***************#
### If turn on the MCP server:
nacos.istio.mcp.server.enabled=false
#*************** Core Related Configurations ***************#
### set the WorkerID manually
# nacos.core.snowflake.worker-id=
### Member-MetaData
# nacos.core.member.meta.site=
# nacos.core.member.meta.adweight=
# nacos.core.member.meta.weight=
### MemberLookup
### Addressing pattern category, If set, the priority is highest
# nacos.core.member.lookup.type=[file,address-server]
## Set the cluster list with a configuration file or command-line argument
# nacos.member.list=192.168.16.101:8847?raft_port=8807,192.168.16.101?raft_port=8808,192.168.16.101:8849?raft_port=8809
## for AddressServerMemberLookup
# Maximum number of retries to query the address server upon initialization
# nacos.core.address-server.retry=5
## Server domain name address of [address-server] mode
# address.server.domain=jmenv.tbsite.net
## Server port of [address-server] mode
# address.server.port=8080
## Request address of [address-server] mode
# address.server.url=/nacos/serverlist
#*************** JRaft Related Configurations ***************#
### Sets the Raft cluster election timeout, default value is 5 second
# nacos.core.protocol.raft.data.election_timeout_ms=5000
### Sets the amount of time the Raft snapshot will execute periodically, default is 30 minute
# nacos.core.protocol.raft.data.snapshot_interval_secs=30
### raft internal worker threads
# nacos.core.protocol.raft.data.core_thread_num=8
### Number of threads required for raft business request processing
# nacos.core.protocol.raft.data.cli_service_thread_num=4
### raft linear read strategy. Safe linear reads are used by default, that is, the Leader tenure is confirmed by heartbeat
# nacos.core.protocol.raft.data.read_index_type=ReadOnlySafe
### rpc request timeout, default 5 seconds
# nacos.core.protocol.raft.data.rpc_request_timeout_ms=5000
#*************** Distro Related Configurations ***************#
### Distro data sync delay time, when sync task delayed, task will be merged for same data key. Default 1 second.
# nacos.core.protocol.distro.data.sync.delayMs=1000
### Distro data sync timeout for one sync data, default 3 seconds.
# nacos.core.protocol.distro.data.sync.timeoutMs=3000
### Distro data sync retry delay time when sync data failed or timeout, same behavior with delayMs, default 3 seconds.
# nacos.core.protocol.distro.data.sync.retryDelayMs=3000
### Distro data verify interval time, verify synced data whether expired for a interval. Default 5 seconds.
# nacos.core.protocol.distro.data.verify.intervalMs=5000
### Distro data verify timeout for one verify, default 3 seconds.
# nacos.core.protocol.distro.data.verify.timeoutMs=3000
### Distro data load retry delay when load snapshot data failed, default 30 seconds.
# nacos.core.protocol.distro.data.load.retryDelayMs=30000
-
初始化mysql数据库，创建数据库表，数据库初始化文件：mysql-schema.sql
(mysql版本要求：5.6.5+) -
单机启动命令：
startup.cmd -m standalone
- 启动报错可能原因分析：
数据库没配置好，库表没创建好，或者数据库名和url后的库名不一致
data目录是之前测试时用过的，可能冲突了，测试环境下直接把data目录删除
nacos客户端配置
server:
port: 9910
spring:
datasource:
# MySql
url: jdbc:mysql://127.0.0.1:3306/test?useSSL=false&useUnicode=true&characterEncoding=utf-8&zeroDateTimeBehavior=convertToNull&transformedBitIsBoolean=true&serverTimezone=GMT%2B8&nullCatalogMeansCurrent=true&allowPublicKeyRetrieval=true
username: root
password: root
driver-class-name: com.mysql.cj.jdbc.Driver
application:
name: file
cloud:
nacos:
config:
import-check:
enabled: false
discovery:
# server in local can only be 127.0.0.1 but not localhost or ip
server-addr: 127.0.0.1:8848
group: SEATA_GROUP
seata服务端配置
- 创建seata\script\server\db sql脚本中的库表
(如果按照博主的教程 需自行创建test库) - 修改seata\conf配置
下面配置中，注释了的 都是非常要小心的 因为客户端以及nacos控制台配置中，需要相对应！
server:
port: 7091
spring:
application:
# seata 服务端名字 , 客户端配置时 需要一致
name: seata-server
logging:
config: classpath:logback-spring.xml
file:
path: ${user.home}/logs/seata
extend:
logstash-appender:
destination: 127.0.0.1:4560
kafka-appender:
bootstrap-servers: 127.0.0.1:9092
topic: logback_to_logstash
console:
user:
username: seata
password: seata
seata:
config:
# support: nacos, consul, apollo, zk, etcd3
type: nacos
nacos:
# nacos 地址
server-addr: 127.0.0.1:8848
# nacos中需要创建该namespace
namespace: fbb5b0a8-7b5f-413e-80b9-4010646e6378
# seata server在nacos的组 客户端配置时 需要一致
group: SEATA_GROUP
username: nacos
password: nacos
# nacos中需要创建该配置
data-id: seataServer.properties
registry:
# support: nacos, eureka, redis, zk, consul, etcd3, sofa
# 类型指定为nacos， 下面
type: nacos
nacos:
# 与application name保持一致
application: seata-server
server-addr: 127.0.0.1:8848
# nacos中需要创建该namespace
namespace: fbb5b0a8-7b5f-413e-80b9-4010646e6378
# seata server在nacos的组 客户端配置时 需要一致
group: SEATA_GROUP
username: nacos
password: nacos
store:
# support: file 、 db 、 redis
# 设置为db模式 记得要创建数据库表
mode: db
db:
datasource: druid
dbType: mysql
driverClassName: com.mysql.cj.jdbc.Driver
url: jdbc:mysql://127.0.0.1:3306/test?useSSL=false&useUnicode=true&characterEncoding=utf-8&zeroDateTimeBehavior=convertToNull&transformedBitIsBoolean=true&serverTimezone=GMT%2B8&nullCatalogMeansCurrent=true&allowPublicKeyRetrieval=true
user: root
password: root
server:
service-port: 8091 #If not configured, the default is '${server.port} + 1000'
security:
secretKey: SeataSecretKey0c382ef121d778043159209298fd40bf3850a017
tokenValidityInMilliseconds: 1800000
ignore:
urls: /,/**/*.css,/**/*.js,/**/*.html,/**/*.map,/**/*.svg,/**/*.png,/**/*.jpeg,/**/*.ico,/api/v1/auth/login
-
启动seata
启动报错可能原因：
a. 数据库配置错误
b. nacos中创建过或使用过seata提供的脚本导入至nacos配置，没有删除
（例如nacos中旧配置指向数据库aaa, 现在我们在seata的application配置文件中改为数据库bbb, 是会优先读取nacos里面的配置的，这个时候如果我们把数据库aaa删除了，启动seata就会报错aaa.xxx table not exist, 我们只要在nacos中将旧配置删除即可)c. 没有seata/lib/jdbc目录下多余的jar包删除, 例如我们mysql版本为8 , 需要将lib包下的mysql5的驱动给手动删掉
(博主测试发现 seata1.6需要手动删除, seata1.7不需要)
nacos中创建seata相关配置
-
在nacos中创建 namespace和data-id为seataServer.properties的配置
【警告】本步骤非常重要，不想部署一次调一次源码，最好不要节外生枝，有些版本的seata，即使客户端中yml配置正确，节点也没写错，但是不在nacos中配置是会报错的, 配置文件错误报的错 包括不限于:can not get cluster name in registry config
no available service
seata :Failed to get available servers:错误可能出现在启动时，也可能出现在项目运行后控制台报错，也可能在java代码调用时抛异常，这取决于我们配置错的有多离谱（后文还会提到客户端配置不当 也可能导致该错误）
a. 创建namespace
如果是按博主教程来的，这儿的id填写 (注意有空格需要去空格)
fbb5b0a8-7b5f-413e-80b9-4010646e6378
b.切换命名空间 创建data id为seataServer.properties的配置
（配置模板在seata服务端目录 seata\script\config-center的config.txt下 ，我们可以先在config中修改好，再复制到naocs）
特别需要注意 service.vgroupMapping.default_tx_group=default 这个配置
(default_tx_group和default 是可以修改的，我知道同学你想改 但是现在先别改，等熟练之后再改，不然很容易出错)
#For details about configuration items, see https://seata.io/zh-cn/docs/user/configurations.html
#Transport configuration, for client and server
transport.type=TCP
transport.server=NIO
transport.heartbeat=true
transport.enableTmClientBatchSendRequest=false
transport.enableRmClientBatchSendRequest=true
transport.enableTcServerBatchSendResponse=false
transport.rpcRmRequestTimeout=30000
transport.rpcTmRequestTimeout=30000
transport.rpcTcRequestTimeout=30000
transport.threadFactory.bossThreadPrefix=NettyBoss
transport.threadFactory.workerThreadPrefix=NettyServerNIOWorker
transport.threadFactory.serverExecutorThreadPrefix=NettyServerBizHandler
transport.threadFactory.shareBossWorker=false
transport.threadFactory.clientSelectorThreadPrefix=NettyClientSelector
transport.threadFactory.clientSelectorThreadSize=1
transport.threadFactory.clientWorkerThreadPrefix=NettyClientWorkerThread
transport.threadFactory.bossThreadSize=1
transport.threadFactory.workerThreadSize=default
transport.shutdown.wait=3
transport.serialization=seata
transport.compressor=none
#Transaction routing rules configuration, only for the client
service.vgroupMapping.default_tx_group=default
#If you use a registry, you can ignore it
service.default.grouplist=127.0.0.1:8091
service.enableDegrade=false
service.disableGlobalTransaction=false
#Transaction rule configuration, only for the client
client.rm.asyncCommitBufferLimit=10000
client.rm.lock.retryInterval=10
client.rm.lock.retryTimes=30
client.rm.lock.retryPolicyBranchRollbackOnConflict=true
client.rm.reportRetryCount=5
client.rm.tableMetaCheckEnable=true
client.rm.tableMetaCheckerInterval=60000
client.rm.sqlParserType=druid
client.rm.reportSuccessEnable=false
client.rm.sagaBranchRegisterEnable=false
client.rm.sagaJsonParser=fastjson
client.rm.tccActionInterceptorOrder=-2147482648
client.tm.commitRetryCount=5
client.tm.rollbackRetryCount=5
client.tm.defaultGlobalTransactionTimeout=60000
client.tm.degradeCheck=false
client.tm.degradeCheckAllowTimes=10
client.tm.degradeCheckPeriod=2000
client.tm.interceptorOrder=-2147482648
client.undo.dataValidation=true
client.undo.logSerialization=jackson
client.undo.onlyCareUpdateColumns=true
server.undo.logSaveDays=7
server.undo.logDeletePeriod=86400000
client.undo.logTable=undo_log
client.undo.compress.enable=true
client.undo.compress.type=zip
client.undo.compress.threshold=64k
#For TCC transaction mode
tcc.fence.logTableName=tcc_fence_log
tcc.fence.cleanPeriod=1h
#Log rule configuration, for client and server
log.exceptionRate=100
#Transaction storage configuration, only for the server. The file, db, and redis configuration values are optional.
store.mode=db
store.lock.mode=db
store.session.mode=db
#Used for password encryption
#store.publicKey=
#If `store.mode,store.lock.mode,store.session.mode` are not equal to `file`, you can remove the configuration block.
#store.file.dir=file_store/data
#store.file.maxBranchSessionSize=16384
#store.file.maxGlobalSessionSize=512
#store.file.fileWriteBufferCacheSize=16384
#store.file.flushDiskMode=async
#store.file.sessionReloadReadSize=100
#These configurations are required if the `store mode` is `db`. If `store.mode,store.lock.mode,store.session.mode` are not equal to `db`, you can remove the configuration block.
store.db.datasource=druid
store.db.dbType=mysql
store.db.driverClassName=com.mysql.cj.jdbc.Driver
store.db.url=jdbc:mysql://127.0.0.1:3306/test?useSSL=false&useUnicode=true&characterEncoding=utf-8&zeroDateTimeBehavior=convertToNull&transformedBitIsBoolean=true&serverTimezone=GMT%2B8&nullCatalogMeansCurrent=true&allowPublicKeyRetrieval=true
store.db.user=root
store.db.password=root
store.db.minConn=5
store.db.maxConn=30
store.db.globalTable=global_table
store.db.branchTable=branch_table
store.db.distributedLockTable=distributed_lock
store.db.queryLimit=100
store.db.lockTable=lock_table
store.db.maxWait=5000
#These configurations are required if the `store mode` is `redis`. If `store.mode,store.lock.mode,store.session.mode` are not equal to `redis`, you can remove the configuration block.
#store.redis.mode=single
#store.redis.single.host=127.0.0.1
#store.redis.single.port=6379
#store.redis.sentinel.masterName=
#store.redis.sentinel.sentinelHosts=
#store.redis.maxConn=10
#store.redis.minConn=1
#store.redis.maxTotal=100
#store.redis.database=0
#store.redis.password=
#store.redis.queryLimit=100
#Transaction rule configuration, only for the server
server.recovery.committingRetryPeriod=1000
server.recovery.asynCommittingRetryPeriod=1000
server.recovery.rollbackingRetryPeriod=1000
server.recovery.timeoutRetryPeriod=1000
server.maxCommitRetryTimeout=-1
server.maxRollbackRetryTimeout=-1
server.rollbackRetryTimeoutUnlockEnable=false
server.distributedLockExpireTime=10000
server.xaerNotaRetryTimeout=60000
server.session.branchAsyncQueueSize=5000
server.session.enableBranchAsyncRemove=false
server.enableParallelRequestHandle=false
#Metrics configuration, only for the server
metrics.enabled=false
metrics.registryType=compact
metrics.exporterList=prometheus
metrics.exporterPrometheusPort=9898
c. nacos中创建data id 为 service.vgroupMapping.default_tx_group 的配置
(这也是为什么上一步说要特别注意这个配置，只在客户端配置这个是不生效的)
此时我们先检查seata服务端是否接入nacos,如果没问题再往下(记得切换命名空间)
seata客户端配置
nacos整合seata之后的完整配置
server:
port: 9910
spring:
datasource:
# MySql
url: jdbc:mysql://127.0.0.1:3306/test?useSSL=false&useUnicode=true&characterEncoding=utf-8&zeroDateTimeBehavior=convertToNull&transformedBitIsBoolean=true&serverTimezone=GMT%2B8&nullCatalogMeansCurrent=true&allowPublicKeyRetrieval=true
username: root
password: root
driver-class-name: com.mysql.cj.jdbc.Driver
application:
name: file
cloud:
nacos:
config:
import-check:
enabled: false
discovery:
# server in local can only be 127.0.0.1 but not localhost or ip
server-addr: 127.0.0.1:8848
# nacos控制台中 我们看到的分组
group: SEATA_GROUP
# nacos控制台中 我们看到的namespace
namespace: fbb5b0a8-7b5f-413e-80b9-4010646e6378
feign:
# 降级： 替代旧版本的 feign.hystrix.enabled配置 (开启会导致seata失效 )
circuitbreaker:
enabled: true
# Seata 配置
seata:
application-id: seata-server
# 这个配置节点的value熟悉吗 是的 前文提醒过要注意的地方了
tx-service-group: default_tx_group
# 是否启用数据源bean的自动代理 默认AT模式，依赖seata-starter时 自动代理即可
# 非常重要：AT（默认）模式下，必须要有代理才能完成全局回滚！！
# 如果为false 必须手动创建代理对象！
enable-auto-data-source-proxy: true
registry:
type: nacos
nacos:
# Nacos 服务地址
server-addr: 127.0.0.1:8848
group: SEATA_GROUP
application: seata-server # 必须和服务端保持一致
username: nacos
password: nacos
cluster: default
namespace: fbb5b0a8-7b5f-413e-80b9-4010646e6378
config:
type: nacos
nacos:
server-addr: ${spring.cloud.nacos.discovery.server-addr}
# namespace: ${spring.cloud.nacos.discovery.namespace}
group: SEATA_GROUP
data-id: seataServer.properties
namespace: fbb5b0a8-7b5f-413e-80b9-4010646e6378
service:
vgroup-mapping:
# 这个key value熟悉吧(虽然客户端配置不生效)是的这个配置有点怪, 把用户可以自定义的值作为key
default_tx_group: default
disable-global-transaction: false
client:
rm:
# 是否上报成功状态
report-success-enable: true
# 重试次数
report-retry-count: 5
mybatis-plus:
mapper-locations: classpath*:mapper/*.xml
logging:
level:
root: info
项目中使用seata方式补充
- 创建我们的业务表,创建java项目
- 还需要创建undo_log表
CREATE TABLE `undo_log` (
`id` bigint NOT NULL AUTO_INCREMENT,
`branch_id` bigint NOT NULL,
`xid` varchar(100) NOT NULL,
`context` varchar(128) NOT NULL,
`rollback_info` longblob NOT NULL,
`log_status` int NOT NULL,
`log_created` datetime NOT NULL,
`log_modified` datetime NOT NULL,
`ext` varchar(100) DEFAULT NULL,
PRIMARY KEY (`id`),
UNIQUE KEY `ux_undo_log` (`xid`,`branch_id`)
) ENGINE=InnoDB AUTO_INCREMENT=8 DEFAULT CHARSET=utf8mb3;
-
业务方法前加上 @GlobalTransactional
-
pom文件示例(仅做参考 )
父模块：
<?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 https://maven.apache.org/xsd/maven-4.0.0.xsd">
<modelVersion>4.0.0</modelVersion>
<parent>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-parent</artifactId>
<version>2.6.13</version>
<relativePath/> <!-- lookup parent from repository -->
</parent>
<!-- 子项目继承父项目pom文件时 父项目需要packing指定为pom
daclare the packing:pom when child extends this moudule-->
<packaging>pom</packaging>
<groupId>com.demo</groupId>
<artifactId>MySpringCloud</artifactId>
<version>0.0.1-SNAPSHOT</version>
<name>MySpringCloud</name>
<description>Demo project for Spring Boot</description>
<!-- 配置版本 version config 用于依赖管理版本号el表达式取值
act on dependencyManagement-dependencies-dependency-version by ${xxx.version}-->
<properties>
<java.version>11</java.version>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
<spring-cloud.version>2021.0.5</spring-cloud.version>
<spring-cloud-alibaba.version>2021.0.5.0</spring-cloud-alibaba.version>
<nacos.client.version>2.2.0</nacos.client.version>
<sentinel.version>1.8.6</sentinel.version>
<netflix-hystrix.version>2.2.10.RELEASE</netflix-hystrix.version>
<mybatis-plus.version>3.5.1</mybatis-plus.version>
<fastjson.version>1.2.76</fastjson.version>
<servlet.version>4.0.1</servlet.version>
<skywalking.version>8.5.0</skywalking.version>
<seata.version>1.6.1</seata.version>
<mysql.version>8.0.32</mysql.version>
<lombok.version>1.18.12</lombok.version>
</properties>
<!-- author by qkj-->
<!-- 声明依赖及其版本 但不引入, 子模块进行引入 （但子模块无需声明版本 否则会使用声明的版本号）-->
<!-- note: dependencyManagement only declare the dependency and version but not import,
child moudule should import again without declare the version when need dependency -->
<!-- 注意：dependencyManagement与dependencies区别 如果在父模块使用dependencies，子模块都继承，不管用不用得上-->
<!-- note: difference with dependencies:child moudule extends dependencies unconditional -->
<dependencyManagement>
<dependencies>
<!-- diff:cloud-dependencies and cloud-starter?
dependencies only declare the version and download jar with starter -->
<!-- cloud-dependencies and cloud-starter 的区别 一般带 dependencies 的 artifactId 都是声明版本
点进去看就能发现 里面是声明的dependencyManagement， 而starter则是我们要下载具体的xxx-starter.jar包-->
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-dependencies</artifactId>
<version>${spring-cloud.version}</version>
<!-- 需要加上type 否则在阿里云可能下载不了
declare type:pom or maybe download fall in alibaba maven
type pom: 已打包为pom-->
<type>pom</type>
<scope>import</scope>
</dependency>
<!-- alibaba与cloud整合 -->
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-alibaba-dependencies</artifactId>
<version>${spring-cloud-alibaba.version}</version>
<type>pom</type>
<!-- scope:import can only declare in dependencyManagement; resolve the problem of single extends -->
<!--<scope>import</scope> 只会在dependencyManagement中出现-->
<!-- 解决单继承问题 换句话说 加上import 可以使用spring-cloud-alibaba-dependencies面的dependencyManagement -->
<!-- such as spring-cloud-alibaba-dependencies already include spring-cloud-starter-alibaba-nacos-discovery
child module can declare spring-cloud-starter-alibaba-nacos-discovery without version-->
<scope>import</scope>
</dependency>
<!--lombok-->
<dependency>
<groupId>org.projectlombok</groupId>
<artifactId>lombok</artifactId>
<optional>true</optional>
<version>${lombok.version}</version>
</dependency>
<!--fastjson 注意过低版本会有漏洞(建议关注新版) low version such as 1.2.71 before with remote bug-->
<dependency>
<groupId>com.alibaba</groupId>
<artifactId>fastjson</artifactId>
<version>${fastjson.version}</version>
</dependency>
<!-- nacos注册中心客户端 内置ribbon-->
<dependency>
<groupId>com.alibaba.nacos</groupId>
<artifactId>nacos-client</artifactId>
<version>${nacos.client.version}</version>
</dependency>
<!-- sentinel-->
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-core</artifactId>
<version>${sentinel.version}</version>
</dependency>
<!-- sentinel控制台-->
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-transport-simple-http</artifactId>
<version>${sentinel.version}</version>
</dependency>
<!-- feign集成hystrix豪猪降级-->
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-starter-netflix-hystrix</artifactId>
<version>${netflix-hystrix.version}</version>
</dependency>
<!-- servlet -->
<dependency>
<groupId>javax.servlet</groupId>
<artifactId>javax.servlet-api</artifactId>
<version>${servlet.version}</version>
</dependency>
<!-- skywalking工具类：自定义链路追踪等-->
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-trace</artifactId>
<!-- 与sw版本对应 -->
<version>${skywalking.version}</version>
</dependency>
<!-- skywalking和logback整合-->
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-logback-1.x</artifactId>
<version>${skywalking.version}</version>
</dependency>
<!--mybatis plus-->
<dependency>
<groupId>com.baomidou</groupId>
<artifactId>mybatis-plus-boot-starter</artifactId>
<version>${mybatis-plus.version}</version>
</dependency>
<!-- mybatis plus 动态数据源 -->
<dependency>
<groupId>com.baomidou</groupId>
<artifactId>dynamic-datasource-spring-boot-starter</artifactId>
<version>${mybatis-plus.version}</version>
</dependency>
<!-- seata -->
<dependency>
<groupId>io.seata</groupId>
<artifactId>seata-spring-boot-starter</artifactId>
<version>${seata.version}</version>
</dependency>
<!--解决seata序列化问题(jackson版本冲突问题)-->
<dependency>
<groupId>io.seata</groupId>
<artifactId>seata-serializer-kryo</artifactId>
<version>${seata.version}</version>
</dependency>
<!-- mysql -->
<dependency>
<groupId>com.mysql</groupId>
<artifactId>mysql-connector-j</artifactId>
<version>${mysql.version}</version>
</dependency>
</dependencies>
</dependencyManagement>
<!-- 可以理解为声明子类(子模块) -->
<modules>
<module>common</module>
<module>product</module>
<module>gateway</module>
<module>user</module>
<module>file</module>
</modules>
<!--注意 子模块有工具类（没有main方法） 不能在父模块这里声明maven插件 子模块继承 会导致打包报找不到main方法(unalbe to find main class) -->
</project>
子模块：
<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 https://maven.apache.org/xsd/maven-4.0.0.xsd">
<modelVersion>4.0.0</modelVersion>
<parent>
<groupId>com.demo</groupId>
<artifactId>MySpringCloud</artifactId>
<version>0.0.1-SNAPSHOT</version>
</parent>
<!-- 因为user是我们的业务模块 通俗点说是一个小项目 到时候要打包java -jar运行的 所以打包方式为jar(可以不写 默认为jar) 而不是pom-->
<packaging>jar</packaging>
<groupId>com.qkj</groupId>
<artifactId>user</artifactId>
<version>0.0.1-SNAPSHOT</version>
<name>user</name>
<description>Demo project for Spring Boot</description>
<properties>
<java.version>1.8</java.version>
</properties>
<dependencies>
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-dependencies</artifactId>
<!-- 父模块该依赖为pom 子模块该依赖也尽量定义为pom
父模块该依赖未使用<scope>import</scope>：当前模块如果还有子模块必须声明为pom 没有则随意 -->
<!-- 但是父模块该依赖使用了import 可以理解为继承了该依赖内置的dependencyManagement 必须使用pom 且声明版本号-->
<!-- must declare type:pom when child module exist or parent module dependency declare scope:import;
version should be declared if parent module dependency declare scope:import,else type could be default or other-->
<!-- so shoule better declare the type: pom -->
<version>${spring-cloud.version}</version>
<type>pom</type>
</dependency>
<!-- alibaba与cloud整合 -->
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-alibaba-dependencies</artifactId>
<version>${spring-cloud-alibaba.version}</version>
<type>pom</type>
</dependency>
<!-- SpringBoot Web 来源于cloud-dependencies >> cloud-common >> cloud-starter >> boot-starter -->
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-web</artifactId>
</dependency>
<!--lombok-->
<dependency>
<groupId>org.projectlombok</groupId>
<artifactId>lombok</artifactId>
<optional>true</optional>
</dependency>
<!--fastjson 注意过低版本会有漏洞 low version such as 1.2.71 before with remote bug-->
<dependency>
<groupId>com.alibaba</groupId>
<artifactId>fastjson</artifactId>
<version>1.2.76</version>
</dependency>
<!-- nacos注册中心客户端 内置ribbon-->
<dependency>
<groupId>com.alibaba.nacos</groupId>
<artifactId>nacos-client</artifactId>
</dependency>
<!-- sentinel-->
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-core</artifactId>
</dependency>
<!-- sentinel控制台-->
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-transport-simple-http</artifactId>
</dependency>
<!--feign 依赖-->
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-starter-openfeign</artifactId>
</dependency>
<!-- 载入open feign 要用到-->
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-loadbalancer</artifactId>
</dependency>
<!-- 豪猪熔断 -->
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-starter-netflix-hystrix</artifactId>
</dependency>
<!--Nacos Discovery-->
<!-- 在spring-cloud-alibaba-dependencies里面声明了版本 又因为设置了scope：import 所以在我们自己写的父模块pom文件中 不能显式看到-->
<!-- declare in spring-cloud-alibaba-dependencies-->
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-starter-alibaba-nacos-discovery</artifactId>
</dependency>
<!--实现配置的动态变更-->
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-starter-alibaba-nacos-config</artifactId>
</dependency>
<!-- skywalking工具类：自定义链路追踪等-->
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-trace</artifactId>
</dependency>
<!-- skywalking和logback整合 print traceId in our logs-->
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-logback-1.x</artifactId>
</dependency>
<!-- seata -->
<dependency>
<groupId>io.seata</groupId>
<artifactId>seata-spring-boot-starter</artifactId>
</dependency>
<!--解决seata序列化问题(jackson版本冲突)-->
<dependency>
<groupId>io.seata</groupId>
<artifactId>seata-serializer-kryo</artifactId>
</dependency>
<!-- Seata 与 cloud alibaba整合 -->
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-starter-alibaba-seata</artifactId>
<exclusions>
<exclusion>
<groupId>io.seata</groupId>
<artifactId>seata-spring-boot-starter</artifactId>
</exclusion>
</exclusions>
</dependency>
<!-- validation -->
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-validation</artifactId>
</dependency>
<dependency>
<groupId>com.mysql</groupId>
<artifactId>mysql-connector-j</artifactId>
</dependency>
<dependency>
<groupId>com.baomidou</groupId>
<artifactId>mybatis-plus-boot-starter</artifactId>
</dependency>
<dependency>
<groupId>com.qkj</groupId>
<artifactId>file-api</artifactId>
<version>0.0.1-SNAPSHOT</version>
<scope>compile</scope>
</dependency>
</dependencies>
<build>
<plugins>
<plugin>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-maven-plugin</artifactId>
<configuration>
<excludes>
<exclude>
<groupId>org.projectlombok</groupId>
<artifactId>lombok</artifactId>
</exclude>
</excludes>
</configuration>
<version>2.3.4.RELEASE</version>
</plugin>
</plugins>
<resources>
<resource>
<directory>src/main/java</directory>
<filtering>false</filtering>
<includes>
<include>**/*.xml</include>
</includes>
</resource>
</resources>
</build>
</project>
全局事务失效原因
-
配置出问题，或本文中提到需要在nacos创建的配置 并未创建
can not get cluster name in registry config
（no available service）
服务找不到，自然全局回滚也会失效 -
如项目中 开启降级或全局异常，可能会导致全局事务失败 ，同理 自己catch了异常 事务也不回滚，不管八股文怎么说，降级无非是一个解耦的try catch
-
引入了动态数据源 例如dynamic-datasource-spring-boot-starter
dynamic时 ，客户端(java项目)的yml 配置有点区别，自行百度。 -
检查是否创建undo_log表以及是否开启数据源代理（以AT模式举例 不管是手动还是自动代理，必须要有一个代理数据源，引入了seata-spring-boot-starter 用自动代理即可 上文教程中有写）
-
据说继承了WebMvcConfigurationSupport类 ，也会导致失效，好像是被拦截了，改为实现WebMvcConfigurer （有具体代码的同学 感兴趣可以调试一下)
-
在不同服务之间打印输出 xid ,xid获取方法：
String xid = RootContext.getXID();
如果为null 那毫无疑问会失效，以上原因(但不限于) 都会导致xid为null
开启降级时 seata失效的解决方法
我们上文提到过，xid为null事务回滚是会失败的，因为seata就是通过在请求头里面传递一个全局的xid。
-
开启降级时，因为降级是异步的，会导致xid为null
-
降级其实就是一个异步解耦的try catch 异常被捕获了，自然也不能回滚。
所以我们解决失效问题，需要同时解决以上两点，网上的教程几乎都是只解决了其中一点； -
还有说在降级时抛异常 ，回滚的…
我们来简单分析一下为什么不可行：
如果没有额外代码编写，降级是异步的，RootContext.getXID()怎么可能获取到值呢？自行debug就能发现为null
可能有的教程刚好是这里报错了 中断运行 当做回滚了吧；
此外，在降级里面回滚和抛异常 似乎改变了降级原本的设计用途
Java代码:
首先编写两个传递xid的类，（解决openfeign降级时 xid为空的问题 ）
import feign.RequestInterceptor;
import feign.RequestTemplate;
import io.seata.core.context.RootContext;
import org.springframework.stereotype.Component;
import org.springframework.web.context.request.RequestContextHolder;
import org.springframework.web.context.request.ServletRequestAttributes;
import javax.servlet.http.HttpServletRequest;
import java.util.*;
/**
* feign 熔断器开启之后 传递 token验证header 头使用
*/
@Component
@ConditionalOnProperty(value = "feign.circuitbreaker.enabled", havingValue = "true",matchIfMissing = false)
public class FeignRequestInterceptor implements RequestInterceptor {
@Override
public void apply(RequestTemplate requestTemplate) {
HttpServletRequest request = getHttpServletRequest();
if (Objects.isNull(request)) {
return;
}
Map<String, String> headers = getHeaders(request);
if (headers.size() > 0) {
Iterator<Map.Entry<String, String>> iterator = headers.entrySet().iterator();
while (iterator.hasNext()) {
Map.Entry<String, String> entry = iterator.next();
requestTemplate.header(entry.getKey(), entry.getValue());
}
}
}
private HttpServletRequest getHttpServletRequest() {
try {
// 这种方式获取的HttpServletRequest是线程安全的
return ((ServletRequestAttributes) (RequestContextHolder.currentRequestAttributes())).getRequest();
} catch (Exception e) {
return null;
}
}
private Map<String, String> getHeaders(HttpServletRequest request) {
Map<String, String> map = new LinkedHashMap<>();
Enumeration<String> enums = request.getHeaderNames();
String kid = (String) request.getAttribute(RootContext.KEY_XID);
while (enums.hasMoreElements()) {
String key = enums.nextElement();
String value = request.getHeader(key);
map.put(key, value);
}
if (kid != null && !kid.equals("")) {
map.put(RootContext.KEY_XID, kid);
}
return map;
}
}
import com.netflix.hystrix.HystrixThreadPoolKey;
import com.netflix.hystrix.HystrixThreadPoolProperties;
import com.netflix.hystrix.strategy.HystrixPlugins;
import com.netflix.hystrix.strategy.concurrency.HystrixConcurrencyStrategy;
import com.netflix.hystrix.strategy.concurrency.HystrixRequestVariable;
import com.netflix.hystrix.strategy.concurrency.HystrixRequestVariableLifecycle;
import com.netflix.hystrix.strategy.eventnotifier.HystrixEventNotifier;
import com.netflix.hystrix.strategy.executionhook.HystrixCommandExecutionHook;
import com.netflix.hystrix.strategy.metrics.HystrixMetricsPublisher;
import com.netflix.hystrix.strategy.properties.HystrixPropertiesStrategy;
import com.netflix.hystrix.strategy.properties.HystrixProperty;
import io.seata.core.context.RootContext;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.stereotype.Component;
import org.springframework.web.context.request.RequestAttributes;
import org.springframework.web.context.request.RequestContextHolder;
import java.util.concurrent.BlockingQueue;
import java.util.concurrent.Callable;
import java.util.concurrent.ThreadPoolExecutor;
import java.util.concurrent.TimeUnit;
@Component
@ConditionalOnProperty(value = "feign.circuitbreaker.enabled", havingValue = "true",matchIfMissing = false)
public class RequestAttributeHystrixConcurrencyStrategy extends HystrixConcurrencyStrategy {
private final Logger log = LoggerFactory.getLogger(getClass());
private HystrixConcurrencyStrategy delegate;
public RequestAttributeHystrixConcurrencyStrategy() {
try {
this.delegate = HystrixPlugins.getInstance().getConcurrencyStrategy();
if (this.delegate instanceof RequestAttributeHystrixConcurrencyStrategy) {
// Welcome to singleton hell...
return;
}
HystrixCommandExecutionHook commandExecutionHook = HystrixPlugins
.getInstance().getCommandExecutionHook();
HystrixEventNotifier eventNotifier = HystrixPlugins.getInstance()
.getEventNotifier();
HystrixMetricsPublisher metricsPublisher = HystrixPlugins.getInstance()
.getMetricsPublisher();
HystrixPropertiesStrategy propertiesStrategy = HystrixPlugins.getInstance()
.getPropertiesStrategy();
this.logCurrentStateOfHystrixPlugins(eventNotifier, metricsPublisher,
propertiesStrategy);
HystrixPlugins.reset();
HystrixPlugins.getInstance().registerConcurrencyStrategy(this);
HystrixPlugins.getInstance()
.registerCommandExecutionHook(commandExecutionHook);
HystrixPlugins.getInstance().registerEventNotifier(eventNotifier);
HystrixPlugins.getInstance().registerMetricsPublisher(metricsPublisher);
HystrixPlugins.getInstance().registerPropertiesStrategy(propertiesStrategy);
}
catch (Exception e) {
}
}
private void logCurrentStateOfHystrixPlugins(HystrixEventNotifier eventNotifier,
HystrixMetricsPublisher metricsPublisher,
HystrixPropertiesStrategy propertiesStrategy) {
if (log.isDebugEnabled()) {
log.debug("Current Hystrix plugins configuration is ["
+ "concurrencyStrategy [" + this.delegate + "]," + "eventNotifier ["
+ eventNotifier + "]," + "metricPublisher [" + metricsPublisher + "],"
+ "propertiesStrategy [" + propertiesStrategy + "]," + "]");
log.debug("Registering Sleuth Hystrix Concurrency Strategy.");
}
}
@Override
public <T> Callable<T> wrapCallable(Callable<T> callable) {
RequestAttributes requestAttributes = RequestContextHolder.getRequestAttributes();
String xid = RootContext.getXID();
requestAttributes.setAttribute(RootContext.KEY_XID,xid, RequestAttributes.SCOPE_REQUEST);
return new WrappedCallable<>(callable, requestAttributes);
}
@Override
public ThreadPoolExecutor getThreadPool(HystrixThreadPoolKey threadPoolKey,
HystrixProperty<Integer> corePoolSize,
HystrixProperty<Integer> maximumPoolSize,
HystrixProperty<Integer> keepAliveTime, TimeUnit unit,
BlockingQueue<Runnable> workQueue) {
return this.delegate.getThreadPool(threadPoolKey, corePoolSize, maximumPoolSize,
keepAliveTime, unit, workQueue);
}
@Override
public ThreadPoolExecutor getThreadPool(HystrixThreadPoolKey threadPoolKey,
HystrixThreadPoolProperties threadPoolProperties) {
return this.delegate.getThreadPool(threadPoolKey, threadPoolProperties);
}
@Override
public BlockingQueue<Runnable> getBlockingQueue(int maxQueueSize) {
return this.delegate.getBlockingQueue(maxQueueSize);
}
@Override
public <T> HystrixRequestVariable<T> getRequestVariable(
HystrixRequestVariableLifecycle<T> rv) {
return this.delegate.getRequestVariable(rv);
}
static class WrappedCallable<T> implements Callable<T> {
private final Callable<T> target;
private final RequestAttributes requestAttributes;
public WrappedCallable(Callable<T> target, RequestAttributes requestAttributes) {
this.target = target;
this.requestAttributes = requestAttributes;
}
@Override
public T call() throws Exception {
try {
RequestContextHolder.setRequestAttributes(requestAttributes);
return target.call();
}
finally {
RequestContextHolder.resetRequestAttributes();
}
}
}
}
大致逻辑为：
RequestAttributeHystrixConcurrencyStrategy设置xid至request，FeignRequestInterceptor从request里面获取到xid，并设置到requestTemplate的header里面。
重要：为什么FeignRequestInterceptor里面不能通过RootContext(本质是threadLocal)获取到xid, 我们应该比较容易联想到是不是降级异步线程导致的？于是输出线程名 发现果然不一致。
凡是异步 都可能导致RootContext.getXid()获取不到值，很多教程用这个代码 那是因为他们未开启降级，上游服务至RequestInterceptor (实现类) 都是在同一个线程下。
当没开启降级时，不需要额外的操作 spring-cloud-starter-alibaba-seata 包 通过SPI机制帮我们配置好了
所以我们自己写的这些类 都需要加上条件 仅作用于开启了降级，否则也会导致传递xid至下游失败
**@ConditionalOnProperty(value = "feign.circuitbreaker.enabled", havingValue = "true",matchIfMissing = false**)
以上两个类只是解决了xid为空的问题，那我们还得解决如何回滚的问题：
import io.seata.common.util.StringUtils;
import io.seata.core.context.RootContext;
import io.seata.core.exception.TransactionException;
import io.seata.tm.api.GlobalTransaction;
import io.seata.tm.api.GlobalTransactionContext;
import org.aspectj.lang.JoinPoint;
import org.aspectj.lang.annotation.AfterReturning;
import org.aspectj.lang.annotation.AfterThrowing;
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Before;
import org.aspectj.lang.reflect.MethodSignature;
import org.springframework.stereotype.Component;
import java.lang.reflect.Method;
/**
* @author qkj
*/
@Aspect
@Component
@ConditionalOnProperty(value = "feign.circuitbreaker.enabled", havingValue = "true",matchIfMissing = false)
public class GlobalTransactionAspect {
@Before("execution(* com.qkj.*.service.*.*(..))")
public void before(JoinPoint joinPoint) throws TransactionException {
MethodSignature signature = (MethodSignature) joinPoint.getSignature();
Method method = signature.getMethod();
System.out.println("拦截到需要分布式事务的方法," + method.getName());
// 此处可用redis或者定时任务来获取一个key判断是否需要关闭分布式事务
if (StringUtils.isBlank(RootContext.getXID())) {
GlobalTransaction tx = GlobalTransactionContext.getCurrentOrCreate();
tx.begin(300000, "test-client");
String xid = tx.getXid();
// 模拟动态关闭分布式事务
//if(contains.key(enable=false){ do nothing // 不手动开启即为关闭})
System.out.println("分布式事务id:" + xid);
} else {
System.out.println("分布式事务id:" + RootContext.getXID());
}
}
@AfterThrowing(throwing = "e", pointcut = "execution(* com.qkj.*.service.*.*(..))")
public void doRecoveryActions(Throwable e) throws TransactionException {
System.out.println("方法执行异常: " + e.getMessage());
if (!StringUtils.isBlank(RootContext.getXID())) {
System.out.println("方法执行异常 上游切面获取的xid:---------- " + RootContext.getXID());
GlobalTransactionContext.reload(RootContext.getXID()).rollback();
}
}
@AfterReturning(value = "execution(* com.qkj.*.service.*.*(..))", returning = "result")
public void afterReturning(JoinPoint point, Object result) throws TransactionException {
System.out.println("方法执行结束: " + result);
if ((Boolean) result) {
if (!StringUtils.isBlank(RootContext.getXID())) {
System.out.println("分布式事务Id: " + RootContext.getXID());
GlobalTransactionContext.reload(RootContext.getXID()).commit();
}
}
}
}
这样还是不够， 我们稍微看一下切面代码就可以发现 ，异常时通过执行doRecoveryActions 进行事务回滚，正常结束时通过执行afterReturning进行事务提交, 所以我们关键在于如何让代码出现异常，
首先我们业务代码肯定不能内部去try catch , catch了还怎么回滚呢
网上有些教程是这种状态码去判断，错误示例:
String resCode;
try{
service.doSomething();
resCode = "200";
}catch(Excepton e){
resCode = "500";
}
我们可以用一种些比较巧妙的方式 例如以下是上游service里面的方法:
(下游代码略 代码很容易推出来)
public Boolean addFile() {
String str = "";
File file = new File();
file.setId(1);
file.setName("test");
fileMapper.insert(file);
Boolean insert = userFeignClient.insert();
String xid = RootContext.getXID();
System.out.println("业务方法：上游调用 xid: " + xid + "===============================================");
// 通过openfeign调用，如果发生了异常 Boolean类型的update结果为null(不是boolean)
Boolean update = userFeignClient.update();
str = "11";
// 这里就会出现异常 切面就能够捕获到
return insert && update && StringUtils.isNotBlank(str);
}
至此，一个相对健全的代码编写完毕。
那如果非得在降级里面回滚 我们可以以下方式处理：
（及其不推荐！容易出问题 可以不看）
其实就一行核心代码：GlobalTransactionContext.reload(RootContext.getXID()).rollback();
其它都是模拟的业务代码 不是完整的。
但是要注意GlobalTransactionContext.reload(RootContext.getXID()).rollback();需要放在方法最后，
且需要注意 ，我们以下是通过回调的方式
（在feign降级里面按照通常的方式导包 很可能会有Maven循环依赖问题, 因为我们feign相关的maven子模块 一般都是轻量级的， 业务层所在的Maven模块，很可能已经导入过了feign相关模块，如果在feign又导入业务层 就会循环）
而且function里面 不要再传com..service. 底下的方法，因为我们在切面中开启的事务扫描了service包，切面是异步，降级也是异步，降级里面回调的方法如果还走切面 很可能会导致获取锁失败！
如果听不懂博主在说什么，那么就按照在切面的时候来处理，不要在降级里面整花活，除非确保降级里面没有数据库相关的操作 ，比如只是些 utils处理
@Component
public class UserFeignFallback implements FallbackFactory<UserFeignClient> {
public static ConcurrentMap<String, Function> fallbackFactoryMap = new ConcurrentHashMap<>();
@Override
public UserFeignClient create(Throwable throwable) {
// 模拟补偿业务
// FileDTO file = new FileDTO();
// file.setId(666);
// file.setName("我是降级");
// Function function = fallbackFactoryMap.get(FallbackConstant.USER_FALLBACK_KEY);
// function.apply(file);
System.out.println("---UserFeignFallback捕获 UserFeign发生异常 实现降级--"+throwable.getMessage()+" --- "+throwable.getCause());
/**
* 已在切面中进行捕获
* {@link com.qkj.file.config.GlobalTransactionAspect#doRecoveryActions(Throwable)}
*
* 以下代码注释， 如果不在切面 doRecoveryActions 方法中捕获,想在这边回滚 则打开本类注释
*/
// 模拟手动在降级中手动回滚 （放在补偿业务下面，防止回滚导致线程直接中断）
// try {
// GlobalTransactionContext.reload(RootContext.getXID()).rollback();
// } catch (TransactionException e) {
// e.printStackTrace();
// }
return null;
}
}
关于如何正确的验证分布式事务
我们写demo时， 要确保自己的例子有代表性 例如正确举例：
// 切面手动开启时 则不用@GlobalTransactional
@GlobalTransactional
@Transactional(rollbackFor = Exception.class)
public Boolean test(){
// 上游服务的方法
fileMapper.insert(file);
// 远程调用方法 成功
Boolean insert = userFeignClient.insert();
// 远程调用方法 在该方法里面制造异常 失败
Boolean update = userFeignClient.update();
return xxx;
}
我们再看看seata-samples/springcloud-nacos-seata 里面 官方的误导性举例：
在这个例子里面 有没有@GlobalTransactional 结果都是一样的(尽管框架源码执行逻辑不一样)，这对初学的同学会造成误导，
上游方法(placeOrder)调用下游方法,下游方法(reduce)报错，
下游方法本身有@Transactional注解，发生了异常 stock表它本身就会回滚(虽然这个例子中 还执行不到操作数据库) ,
上游方法placeOrder中 也有@Transactional注解， stockFeignClient.deduce(commodityCode,count) 这行代码会报错，所以order表也会回滚。
博主的例子，多了个insert的操作，如果我们insert方法写的没问题的前提下，参与了回滚 说明seata的回滚生效了
5226
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
