東川印記

一本東川,笑看爭龍斗虎;寰茫兦者,度橫佰昧人生。

简单学习MyCat

2022年7月4日星期一



接上 MySQL集群部署,读写分离

但是代码调用非常麻烦,写一大堆代码。所以就直接没写。。。。


MyCat前身为阿里Cobar,Cobar被阿里放弃后,被开源爱好者改造成了MyCat。

开源数据库中间件,实现了MySQL协议的数据库中间件服务器。

默认端口8066.

主要用来做

1)数据库读写分离:应用连接MyCat自动实现读写分离;

2)分库分表:水平(横向)切分、垂直(纵向)切分。


水平拆分:为了解决大表问题(如1000万条Orders数据),根据表中某列数据拆分,拆分到不同数据库中Orders表以解决大表带来问题。

如1000万条Order,分为5个库,每个库的Order存200万条。根据id%5=0-4存储到不同的位置中。

基于上面的拆分规则,Order业务至少有5*2主库*2从库=20个数据源。可以用MyCat配置规则算法实现。

垂直拆分:一个数据库有100张表,访问人数增加会形成IO竞争。如1000W人访问订单表,1人访问用户表,因数据库连接上限导致1人用户登录变慢,且所有数据全部存放在一个库中,造成磁盘压力。

为解决表特别多问题,使用垂直拆分,根据不同表的业务关系将表存放到不同库。不同表放在不同库,解决IO竞争问题。

如用户模块、订单模块、商品模块等5个模块,不同业务放在不同的库。

基于上面拆分规则,会有 5个模块*4个水平拆分*2主库*2从库= 80个数据源。同样可以桥接MyCat来根据规则实现。

垂直拆分后一般分组按模块独立开发,基本不会跨模块访问。当需要跨模块时,使用对方模块提供的接口。


拆分优先级:缓存降低数据库读操作 -> 读写分离降低数据库写操作 -> 垂直拆分 -> 水平拆分(分库) -> 水平拆分(分表)。

1,环境搭建

下载 https://github.com/MyCATApache/Mycat-Server/releases

现在最后一个稳定版是1.6.7.5 两年没更新了。。。。

解压

[senrsl@localhost mycat]$ tar zxvf Mycat-server-1.6.7.5-release-20200422133810-linux.tar.gz

[senrsl@localhost mycat]$ ll
总用量 22668
drwxrwxr-x. 6 senrsl senrsl       73 7月   2 06:34 mycat
-rw-r--r--. 1 senrsl senrsl 23209170 7月   2 06:33 Mycat-server-1.6.7.5-release-20200422133810-linux.tar.gz
[senrsl@localhost mycat]$

[senrsl@localhost mycat]$ ll mycat/
总用量 12
drwxrwxr-x. 2 senrsl senrsl  190 7月   2 06:34 bin
drwxrwxr-x. 2 senrsl senrsl    6 4月  15 2020 catlet
drwxrwxr-x. 4 senrsl senrsl 4096 7月   2 06:34 conf
drwxrwxr-x. 2 senrsl senrsl 4096 7月   2 06:34 lib
-rwxrwxr-x. 1 senrsl senrsl  227 4月  22 2020 version.txt
[senrsl@localhost mycat]$

2,配置文件server.xml

server.xml

<?xml version="1.0" encoding="UTF-8"?>
<!-- - - Licensed under the Apache License, Version 2.0 (the "License");
        - you may not use this file except in compliance with the License. - You
        may obtain a copy of the License at - - http://www.apache.org/licenses/LICENSE-2.0
        - - Unless required by applicable law or agreed to in writing, software -
        distributed under the License is distributed on an "AS IS" BASIS, - WITHOUT
        WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. - See the
        License for the specific language governing permissions and - limitations
        under the License. -->
<!DOCTYPE mycat:server SYSTEM "server.dtd">
<mycat:server xmlns:mycat="http://io.mycat/">
        <system>
        <property name="nonePasswordLogin">0</property> <!-- 0为需要密码登陆、1为不需要密码登陆 ,默认为0,设置为1则需要指定默认账户-->
        <property name="ignoreUnknownCommand">0</property><!-- 0遇上没有实现的报文(Unknown command:),就会报错、1为忽略该报文,返回ok报文。
        在某些mysql客户端存在客户端已经登录的时候还会继续发送登录报文,mycat会报错,该设置可以绕过这个错误-->
        <property name="useHandshakeV10">1</property>
    <property name="removeGraveAccent">1</property>
        <property name="useSqlStat">0</property>  <!-- 1为开启实时统计、0为关闭 -->
        <property name="useGlobleTableCheck">0</property>  <!-- 1为开启全加班一致性检测、0为关闭 -->
        <property name="sqlExecuteTimeout">300</property>  <!-- SQL 执行超时 单位:秒-->
                <property name="sequenceHandlerType">1</property>
        <!--<property name="sequnceHandlerPattern">(?:(\s*next\s+value\s+for\s*MYCATSEQ_(\w+))(,|\)|\s)*)+</property>
        INSERT INTO `travelrecord` (`id`,user_id) VALUES ('next value for MYCATSEQ_GLOBAL',"xxx");
        -->
        <!--必须带有MYCATSEQ_或者 mycatseq_进入序列匹配流程 注意MYCATSEQ_有空格的情况-->
        <property name="sequnceHandlerPattern">(?:(\s*next\s+value\s+for\s*MYCATSEQ_(\w+))(,|\)|\s)*)+</property>
        <property name="subqueryRelationshipCheck">false</property> <!-- 子查询中存在关联查询的情况下,检查关联字段中是否有分片字段 .默认 false -->
        <property name="sequenceHanlderClass">io.mycat.route.sequence.handler.HttpIncrSequenceHandler</property>
      <!--  <property name="useCompression">1</property>--> <!--1为开启mysql压缩协议-->
        <!--  <property name="fakeMySQLVersion">5.6.20</property>--> <!--设置模拟的MySQL版本号-->
        <!-- <property name="processorBufferChunk">40960</property> -->
        <!--
        <property name="processors">1</property>
        <property name="processorExecutor">32</property>
         -->
        <!--默认为type 0: DirectByteBufferPool | type 1 ByteBufferArena | type 2 NettyBufferPool -->
                <property name="processorBufferPoolType">0</property>
                <!--默认是65535 64K 用于sql解析时最大文本长度 -->
                <!--<property name="maxStringLiteralLength">65535</property>-->
                <!--<property name="sequenceHandlerType">0</property>-->
                <!--<property name="backSocketNoDelay">1</property>-->
                <!--<property name="frontSocketNoDelay">1</property>-->
                <!--<property name="processorExecutor">16</property>-->
                <!--
                        <property name="serverPort">8066</property> <property name="managerPort">9066</property>
                        <property name="idleTimeout">300000</property> <property name="bindIp">0.0.0.0</property>
                        <property name="dataNodeIdleCheckPeriod">300000</property> 5 * 60 * 1000L; //连接空闲检查
                        <property name="frontWriteQueueSize">4096</property> <property name="processors">32</property> -->
                <!--分布式事务开关,0为不过滤分布式事务,1为过滤分布式事务(如果分布式事务内只涉及全局表,则不过滤),2为不过滤分布式事务,但是记录分布式事务日志-->
                <property name="handleDistributedTransactions">0</property>

                        <!--
                        off heap for merge/order/group/limit      1开启   0关闭
                -->
                <property name="useOffHeapForMerge">0</property>

                <!--
                        单位为m
                -->
        <property name="memoryPageSize">64k</property>

                <!--
                        单位为k
                -->
                <property name="spillsFileBufferSize">1k</property>

                <property name="useStreamOutput">0</property>

                <!--
                        单位为m
                -->
                <property name="systemReserveMemorySize">384m</property>


                <!--是否采用zookeeper协调切换  -->
                <property name="useZKSwitch">false</property>

                <!-- XA Recovery Log日志路径 -->
                <!--<property name="XARecoveryLogBaseDir">./</property>-->

                <!-- XA Recovery Log日志名称 -->
                <!--<property name="XARecoveryLogBaseName">tmlog</property>-->
                <!--如果为 true的话 严格遵守隔离级别,不会在仅仅只有select语句的时候在事务中切换连接-->
                <property name="strictTxIsolation">false</property>
                <!--如果为0的话,涉及多个DataNode的catlet任务不会跨线程执行-->
                <property name="parallExecute">0</property>
        </system>

        <!-- 全局SQL防火墙设置 -->
        <!--白名单可以使用通配符%或着*-->
        <!--例如<host host="127.0.0.*" user="root"/>-->
        <!--例如<host host="127.0.*" user="root"/>-->
        <!--例如<host host="127.*" user="root"/>-->
        <!--例如<host host="1*7.*" user="root"/>-->
        <!--这些配置情况下对于127.0.0.1都能以root账户登录-->
        <!--
        <firewall>
           <whitehost>
              <host host="1*7.0.0.*" user="root"/>
           </whitehost>
       <blacklist check="false">
       </blacklist>
        </firewall>
        -->

        <user name="root" defaultAccount="true"> //这里root改成mycat方便与mysql区分
                <property name="password">123456</property>
                <property name="schemas">TESTDB</property>  //两个TESTDB改成MYCATDB,方便区分
                <property name="defaultSchema">TESTDB</property>
                <!--No MyCAT Database selected 错误前会尝试使用该schema作为schema,不设置则为null,报错 -->

                <!-- 表级 DML 权限设置 -->
                <!--
                <privileges check="false">
                        <schema name="TESTDB" dml="0110" >
                                <table name="tb01" dml="0000"></table>
                                <table name="tb02" dml="1111"></table>
                        </schema>
                </privileges>
                 -->
        </user>

        <user name="user">
                <property name="password">user</property>
                <property name="schemas">TESTDB</property>//MYCATDB
                <property name="readOnly">true</property>
                <property name="defaultSchema">TESTDB</property>
        </user>

</mycat:server>

3,配置文件scheme.xml

[senrsl@localhost mycat]$ cat conf/schema.xml
<?xml version="1.0"?>
<!DOCTYPE mycat:schema SYSTEM "schema.dtd">
<mycat:schema xmlns:mycat="http://io.mycat/">

        <schema name="TESTDB" checkSQLschema="true" sqlMaxLimit="100" randomDataNode="dn1">  //逻辑库TESTDB
                <!-- auto sharding by id (long) -->
                <!--splitTableNames 启用<table name 属性使用逗号分割配置多个表,即多个表使用这个配置-->
<!--fetchStoreNodeByJdbc 启用ER表使用JDBC方式获取DataNode-->
                <table name="customer" primaryKey="id" dataNode="dn1,dn2" rule="sharding-by-intfile" autoIncrement="true" fetchStoreNodeByJdbc="true">
                        <childTable name="customer_addr" primaryKey="id" joinKey="customer_id" parentKey="id"> </childTable>
                </table>
                <!-- <table name="oc_call" primaryKey="ID" dataNode="dn1$0-743" rule="latest-month-calldate"
                        /> -->
        </schema>
        <!-- <dataNode name="dn1$0-743" dataHost="localhost1" database="db$0-743"
                /> -->
        <dataNode name="dn1" dataHost="localhost1" database="db1" />
        <dataNode name="dn2" dataHost="localhost1" database="db2" />
        <dataNode name="dn3" dataHost="localhost1" database="db3" />
        <!--<dataNode name="dn4" dataHost="sequoiadb1" database="SAMPLE" />
         <dataNode name="jdbc_dn1" dataHost="jdbchost" database="db1" />
        <dataNode       name="jdbc_dn2" dataHost="jdbchost" database="db2" />
        <dataNode name="jdbc_dn3"       dataHost="jdbchost" database="db3" /> -->
        <dataHost name="localhost1" maxCon="1000" minCon="10" balance="0"
                          writeType="0" dbType="mysql" dbDriver="jdbc" switchType="1"  slaveThreshold="100">
                <heartbeat>select user()</heartbeat>
                <!-- can have multi write hosts -->
                <writeHost host="hostM1" url="jdbc:mysql://localhost:3306" user="root"
                                   password="root">
                </writeHost>
                <!-- <writeHost host="hostM2" url="localhost:3316" user="root" password="123456"/> -->
        </dataHost>
        <!--
                <dataHost name="sequoiadb1" maxCon="1000" minCon="1" balance="0" dbType="sequoiadb" dbDriver="jdbc">
                <heartbeat>             </heartbeat>
                 <writeHost host="hostM1" url="sequoiadb://1426587161.dbaas.sequoialab.net:11920/SAMPLE" user="jifeng"  password="jifeng"></writeHost>
                 </dataHost>

          <dataHost name="oracle1" maxCon="1000" minCon="1" balance="0" writeType="0"   dbType="oracle" dbDriver="jdbc"> <heartbeat>select 1 from dual</heartbeat>
                <connectionInitSql>alter session set nls_date_format='yyyy-mm-dd hh24:mi:ss'</connectionInitSql>
                <writeHost host="hostM1" url="jdbc:oracle:thin:@127.0.0.1:1521:nange" user="base"       password="123456" > </writeHost> </dataHost>

                <dataHost name="jdbchost" maxCon="1000"         minCon="1" balance="0" writeType="0" dbType="mongodb" dbDriver="jdbc">
                <heartbeat>select       user()</heartbeat>
                <writeHost host="hostM" url="mongodb://192.168.0.99/test" user="admin" password="123456" ></writeHost> </dataHost>

                <dataHost name="sparksql" maxCon="1000" minCon="1" balance="0" dbType="spark" dbDriver="jdbc">
                <heartbeat> </heartbeat>
                 <writeHost host="hostM1" url="jdbc:hive2://feng01:10000" user="jifeng"         password="jifeng"></writeHost> </dataHost> -->

        <!-- <dataHost name="jdbchost" maxCon="1000" minCon="10" balance="0" dbType="mysql"
                dbDriver="jdbc"> <heartbeat>select user()</heartbeat> <writeHost host="hostM1"
                url="jdbc:mysql://localhost:3306" user="root" password="123456"> </writeHost>
                </dataHost> -->
</mycat:schema>[senrsl@localhost mycat]$

4,读写分离配置

server.xml 改下面两处

    <user name="mycat" defaultAccount="true">
        <property name="password">123456</property>
        <property name="schemas">MYCATDB</property>
        <property name="defaultSchema">MYCATDB</property>
        <!--No MyCAT Database selected 错误前会尝试使用该schema作为schema,不设置则为null,报错 -->
       
        <!-- 表级 DML 权限设置 -->
        <!--        
        <privileges check="false">
            <schema name="TESTDB" dml="0110" >
                <table name="tb01" dml="0000"></table>
                <table name="tb02" dml="1111"></table>
            </schema>
        </privileges>       
         -->
    </user>

    <user name="user">
        <property name="password">user</property>
        <property name="schemas">MYCATDB</property>
        <property name="readOnly">true</property>
        <property name="defaultSchema">MYCATDB</property>
    </user>

scheme.xml内容如下

<?xml version="1.0"?>
<!DOCTYPE mycat:schema SYSTEM "schema.dtd">
<mycat:schema xmlns:mycat="http://io.mycat/">

    <schema name="MYCATDB" checkSQLschema="true" sqlMaxLimit="100" dataNode="dn1">
       
    </schema>
   
    <dataNode name="dn1" dataHost="localhost1" database="test202201" />
   
    <dataHost name="localhost1" maxCon="1000" minCon="10" balance="1"
              writeType="0" dbType="mysql" dbDriver="native" switchType="1"  slaveThreshold="100">
        <heartbeat>select user()</heartbeat>
        <writeHost host="hostM1" url=" 192.168.0.122:3306" user="root"
                   password="0123456">
            <readHost host="hostS1" url="192.168.0.125:3306" user="root" password="0123456" />
        </writeHost>
        <writeHost host="hostM2" url=" 192.168.0.142:3306" user="root"
                   password="0123456">
            <readHost host="hostS2" url="192.168.0.111:3306" user="root" password="0123456" />
        </writeHost>
    </dataHost>
   
</mycat:schema>


5,启动

启动后查看进程是否存在,否则就去看 log报错。。。。

[senrsl@localhost mycat]$ bin/mycat start
Starting Mycat-server...
[senrsl@localhost mycat]$ ps -ef | grep mycat

报错1:

[senrsl@localhost mycat]$ cat wrapper.log
FATAL  | wrapper  | 2022/07/02 07:15:48 | ERROR: Could not write pid file /home/senrsl/mycat/mycat/logs/mycat.pid: No such file or directory

手动创建logs/mycat.pid

6,设置mysql允许远程连接

mysql> grant all on *.* to 'root'@'%' identified by '0123456' with grant option;
Query OK, 0 rows affected, 1 warning (0.00 sec)

昨天能连,今天莫名其妙就连不上了。执行上面sql就好了。。。。

莫名其妙,之前全都不行,四台机器之设置了一台结果其他三台也行了。。。。

7,读写分离测试

连接mycat,使用连接mysql方式,端口8066,mycat/123456

可以看到里面有MYCATDB这个库。里面有user表。

执行插入数据等测试,确认 mysql主从库同步。

INSERT INTO users(name) VALUES("sql插入测试");

自动实现读写分离。

此时双主双从模式、122、142互为主从、125为122的从,111为142的从。

此时如果122宕机,125从也不会参与读了。只有142参与写,111参与读。

恢复122后,122和125会自动同步数据到最新。

8,SpringBoot调用测试

1)新建TestMyCat SpringBoot项目

2)依赖

<!-- 连接 Mybatis -->
<dependency>
    <groupId>org.mybatis.spring.boot</groupId>
    <artifactId>mybatis-spring-boot-starter</artifactId>
    <version>2.1.4</version>
</dependency>

<!-- 连接MySQL -->
<dependency>
    <groupId>mysql</groupId>
    <artifactId>mysql-connector-java</artifactId>
    <scope>runtime</scope>
    <version>5.1.46</version>
</dependency>

3)数据库连接配置,直接配到MyCat上

# 使用MyCat连接
# 数据库驱动:
spring.datasource.driver-class-name=com.mysql.jdbc.Driver
# 数据源名称
#spring.datasource.name=defaultDataSource
# 数据库连接地址
spring.datasource.url=jdbc:mysql://192.168.0.122:8066/MYCATDB
# 数据库用户名&密码:
spring.datasource.username=mycat
spring.datasource.password=123456

4)添加映射自动生成配置

<build>
    <plugins>
        <plugin>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-maven-plugin</artifactId>
        </plugin>
        <plugin>
            <groupId>org.mybatis.generator</groupId>
            <artifactId>mybatis-generator-maven-plugin</artifactId>
            <version>1.3.6</version>
            <configuration>
                <!--配置文件的位置-->
                <configurationFile>GeneratorMapper.xml</configurationFile>
                <verbose>true</verbose>
                <overwrite>true</overwrite>
            </configuration>
        </plugin>
    </plugins>
    <resources>
        <resource>
            <directory>src/main/java</directory>
            <includes>
                <include>**/*.xml</include>
            </includes>
        </resource>

    </resources>
</build>

5)添加映射文件

项目根目录GeneratorMapper.xml

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE generatorConfiguration
        PUBLIC "-//mybatis.org//DTD MyBatis Generator Configuration 1.0//EN"
        "http://mybatis.org/dtd/mybatis-generator-config_1_0.dtd">

<generatorConfiguration>

    <!-- 指定连接数据库的JDBC驱动包所在位置,指定到你本机的完整路径 -->
    <classPathEntry location="/Users/senrsl/.m2/repository/mysql/mysql-connector-java/5.1.46/mysql-connector-java-5.1.46.jar"/>

    <!-- 配置table表信息内容体,targetRuntime指定采用MyBatis3的版本 -->
    <context id="tables" targetRuntime="MyBatis3">

        <!-- 抑制生成注释,由于生成的注释都是英文的,可以不让它生成 -->
        <commentGenerator>
            <property name="suppressAllComments" value="true" />
        </commentGenerator>

        <!-- 配置数据库连接信息 -->
        <jdbcConnection driverClass="com.mysql.jdbc.Driver"
                        connectionURL="jdbc:mysql://192.168.0.122:8066/MYCATDB"
                        userId="mycat"
                        password="123456">
        </jdbcConnection>

        <!-- 生成model类,targetPackage指定model类的包名, targetProject指定生成的model放在eclipse的哪个工程下面-->
        <javaModelGenerator targetPackage="dc.test.model" targetProject="src/main/java">
            <property name="enableSubPackages" value="false" />
            <property name="trimStrings" value="false" />
        </javaModelGenerator>

        <!-- 生成MyBatis的Mapper.xml文件,targetPackage指定mapper.xml文件的包名, targetProject指定生成的mapper.xml放在eclipse的哪个工程下面 -->
        <sqlMapGenerator targetPackage="dc.test.mapper" targetProject="src/main/java">
            <property name="enableSubPackages" value="false" />
        </sqlMapGenerator>

        <!-- 生成MyBatis的Mapper接口类文件,targetPackage指定Mapper接口类的包名, targetProject指定生成的Mapper接口放在eclipse的哪个工程下面 -->
        <javaClientGenerator type="XMLMAPPER" targetPackage="dc.test.mapper" targetProject="src/main/java">
            <property name="enableSubPackages" value="false" />
        </javaClientGenerator>

        <!-- 数据库表名及对应的Java模型类名 -->
        <table tableName="users"
               domainObjectName="Users"
               enableCountByExample="false"
               enableUpdateByExample="false"
               enableDeleteByExample="false"
               enableSelectByExample="false"
               selectByExampleQueryId="false"/>


    </context>

</generatorConfiguration>

6)映射生成

Maven -> mybatis-generator

配图1

生成文件如下,给Mapper添加注解

配图2

7)编写service

@Service
public class UserService {

    @Resource
    private UsersMapper usersMapper;

    public void insert(Users user){
        usersMapper.insert(user);
    }

    public void selectAll(){
        List<Users> list = usersMapper.selectAll();
        for (Users user:list) {
            System.out.println(user.toString());
        }
    }

}

8)运行

@SpringBootApplication
public class TestApplication {

    public static void main(String[] args) {
        ConfigurableApplicationContext context = SpringApplication.run(TestApplication.class, args);
        String[] arr = context.getBeanNamesForAnnotation(org.springframework.stereotype.Service.class);

        System.out.println("迭代" + arr);
        for (String str : arr) {
            System.out.println("is  " + str);
        }

        UserService userService =  context.getBean("userService",UserService.class);

        userService.insert(new Users("张三" + new Date()));

        userService.selectAll();

    }

}


报错1:

org.springframework.beans.factory.NoSuchBeanDefinitionException: No bean named 'UserService' available

把启动类Application放在跟Service平级的目录会取不到,必须要放在外面才能取到。。。。

此处配图3

9,水平拆分测试

1)环境准备

在vm121创建三个库,test202202、03、04

mysql> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sys                |
| test2022           |
| test202201         |
| test202202         |
| test202203         |
| test202204         |
+--------------------+
9 rows in set (0.00 sec)

mysql>

每个库创建表t_orders

/*
 Navicat Premium Data Transfer

 Source Server         : vm121
 Source Server Type    : MySQL
 Source Server Version : 50737
 Source Host           : 192.168.0.122:3306
 Source Schema         : test202202

 Target Server Type    : MySQL
 Target Server Version : 50737
 File Encoding         : 65001

 Date: 04/07/2022 14:45:57
*/

SET NAMES utf8mb4;
SET FOREIGN_KEY_CHECKS = 0;

-- ----------------------------
-- Table structure for t_orders
-- ----------------------------
DROP TABLE IF EXISTS `t_orders`;
CREATE TABLE `t_orders` (
  `id` bigint(20) NOT NULL AUTO_INCREMENT,
  `name` varchar(128) DEFAULT NULL,
  `createTime` datetime DEFAULT NULL ON UPDATE CURRENT_TIMESTAMP,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

SET FOREIGN_KEY_CHECKS = 1;

其他三个从机也就都同步了这三个库了。。。。

修改:id设置为不自增,把自增给MyCat来实现。。。。

2)修改MyCat配置文件

server.xml同读写分离

scheme.xml

<?xml version="1.0"?>
<!DOCTYPE mycat:schema SYSTEM "schema.dtd">
<mycat:schema xmlns:mycat="http://io.mycat/">

    <schema name="MYCATDB" checkSQLschema="true" sqlMaxLimit="100" randomDataNode="dn1">
       
        <table name="t_orders" primaryKey="id" dataNode="dn1,dn2,dn3" rule="mod-long" autoIncrement="true" >
           
        </table>
       
    </schema>
   
    <dataNode name="dn1" dataHost="localhost1" database="test202202" />
    <dataNode name="dn2" dataHost="localhost1" database="test202203" />
    <dataNode name="dn3" dataHost="localhost1" database="test202204" />

    <dataHost name="localhost1" maxCon="1000" minCon="10" balance="1"
              writeType="0" dbType="mysql" dbDriver="native" switchType="1"  slaveThreshold="100">
        <heartbeat>select user()</heartbeat>
       
        <writeHost host="hostM1" url=" 192.168.0.122:3306" user="root"
                   password="0123456">
            <readHost host="hostS1" url="192.168.0.125:3306" user="root" password="0123456" />
        </writeHost>
        <writeHost host="hostM2" url=" 192.168.0.142:3306" user="root"
                   password="0123456">
            <readHost host="hostS2" url="192.168.0.112:3306" user="root" password="0123456" />
        </writeHost>
    </dataHost>
</mycat:schema>

其中table标签下rule配置于conf/rule.xml

3)替换配置重启服务

[root@localhost mycat]# ./bin/mycat stop
Stopping Mycat-server...
Stopped Mycat-server.
[root@localhost mycat]# ./bin/mycat start
Starting Mycat-server...
[root@localhost mycat]#

4)测试

navicat连接mycat库

插入测试

INSERT into t_orders(name)VALUES('张三');


报错1:

INSERT into t_orders(name)VALUES('张三')
> 1003 - mycat sequnce err.io.mycat.config.util.ConfigException: can't find definition for sequence :T_ORDERS
> Time: 0s

1064 - can't find table define in schema TRIGGERS schema:MYCATDB

能读不能写

修改

[root@localhost mycat]# cat conf/sequence_db_conf.properties
#sequence stored in datanode
GLOBAL=dn1
COMPANY=dn1
CUSTOMER=dn1
ORDERS=dn1
T_ORDERS=dn1   //添加需要自增长的表名称
[root@localhost mycat]#

报错2:

INSERT INTO t_orders(name)VALUES("张三")
> 1003 - mycat sequnce err.java.lang.RuntimeException: can't fetch sequnce in db,sequnce :T_ORDERS detail:FUNCTION test202202.mycat_seq_nextval does not exist
> Time: 0s

解决,三个数据库依次执行

#添加表
DROP TABLE IF EXISTS MYCAT_SEQUENCE;
CREATE TABLE MYCAT_SEQUENCE (
NAME VARCHAR (50) NOT NULL,
current_value INT NOT NULL,
increment INT NOT NULL DEFAULT 100,
PRIMARY KEY (NAME)
) ENGINE = INNODB ;


INSERT INTO MYCAT_SEQUENCE(name,current_value,increment) VALUES ('GLOBAL', 100000, 100);

DROP FUNCTION IF EXISTS `mycat_seq_currval`;
DELIMITER ;;
CREATE FUNCTION `mycat_seq_currval`(seq_name VARCHAR(50))
RETURNS varchar(64) CHARSET utf8
    DETERMINISTIC
BEGIN DECLARE retval VARCHAR(64);
        SET retval="-999999999,null"; 
        SELECT concat(CAST(current_value AS CHAR),",",CAST(increment AS CHAR) ) INTO retval
          FROM MYCAT_SEQUENCE WHERE name = seq_name; 
        RETURN retval ;
END
;;
DELIMITER ;

DROP FUNCTION IF EXISTS `mycat_seq_nextval`;
DELIMITER ;;
CREATE FUNCTION `mycat_seq_nextval`(seq_name VARCHAR(50)) RETURNS varchar(64)
 CHARSET utf8
    DETERMINISTIC
BEGIN UPDATE MYCAT_SEQUENCE 
                 SET current_value = current_value + increment
                  WHERE name = seq_name; 
         RETURN mycat_seq_currval(seq_name); 
END
;;
DELIMITER ;


DROP FUNCTION IF EXISTS `mycat_seq_setval`;
DELIMITER ;;
CREATE FUNCTION `mycat_seq_setval`(seq_name VARCHAR(50), value INTEGER)
RETURNS varchar(64) CHARSET utf8
    DETERMINISTIC
BEGIN UPDATE MYCAT_SEQUENCE 
                   SET current_value = value 
                   WHERE name = seq_name; 
         RETURN mycat_seq_currval(seq_name); 
END
;;
DELIMITER ;


报错3:

INSERT INTO t_orders(name)VALUES("张三")
> 1003 - mycat sequnce err.java.lang.NumberFormatException: For input string: "null"
> Time: 0s

mycat获取自增字段有三种方式:从配置文件0、从数据库1、本地时间戳2

对应配置在 server.xml

<property name="sequenceHandlerType">1</property>

默认是1,改成2使用本地时间戳方式。。。。

终于插入成功了。。。。

敢情前面那些都不用改。。。。

插入后,会自动分到不同的数据库中来存储。。。。

5)SpringBoot测试

生成mapper

<!-- 数据库表名及对应的Java模型类名 -->
<!--<table tableName="users"
       domainObjectName="Users"
       enableCountByExample="false"
       enableUpdateByExample="false"
       enableDeleteByExample="false"
       enableSelectByExample="false"
       selectByExampleQueryId="false"/>-->
<table tableName="t_orders"
       domainObjectName="Orders"
       enableCountByExample="false"
       enableUpdateByExample="false"
       enableDeleteByExample="false"
       enableSelectByExample="false"
       selectByExampleQueryId="false"/>

OrdersMapper增加 @Mapper

编写Service

@Service
public class OrderService {

    @Resource
    private OrdersMapper mapper;

    public void add(Orders orders) {
        mapper.insert(orders);
    }

    public void selectAll() {
        List<Orders> list = mapper.selectAll();
        for (Orders order : list) {
            System.out.println(order.toString());
        }
    }

}

运行

@SpringBootApplication
public class TestOrderApplication {

    public static void main(String[] args) {
        ApplicationContext context = SpringApplication.run(TestOrderApplication.class,args);
        OrderService orderService  = (OrderService) context.getBean("orderService");

        orderService.add(new Orders("李四",new Date()));

        orderService.selectAll();
    }

}


报错:

Error attempting to get column 'name' from result set.  Cause: java.sql.SQLException: Cannot convert value '测试' from column 2 to TIMESTAMP.

无语。。。。为啥不能映射timestamp....

10,垂直拆分测试

1)环境搭建

一个表拆三个库

还是用test202202、03、04

分别建 t_order2、t_good、t_users

都有两列 id、name

配图4

2)配置

server.xml同水平拆分

scheme.xml

<?xml version="1.0"?>
<!DOCTYPE mycat:schema SYSTEM "schema.dtd">
<mycat:schema xmlns:mycat="http://io.mycat/">

    <schema name="MYCATDB" checkSQLschema="true" sqlMaxLimit="100" randomDataNode="dn1">
        <table name="t_order2" primaryKey="id" dataNode="dn1"  autoIncrement="true" >
             
        </table>
        <table name="t_goods" primaryKey="id" dataNode="dn2"  autoIncrement="true" >
             
        </table>
        <table name="t_users" primaryKey="id" dataNode="dn3"  autoIncrement="true"  >
             
        </table>
    </schema>
    <!-- <dataNode name="dn1$0-743" dataHost="localhost1" database="db$0-743"
        /> -->
    <dataNode name="dn1" dataHost="localhost1" database="test202202" />
    <dataNode name="dn2" dataHost="localhost1" database="test202203" />
    <dataNode name="dn3" dataHost="localhost1" database="test202204" />
   
    <dataHost name="localhost1" maxCon="1000" minCon="10" balance="1"
              writeType="0" dbType="mysql" dbDriver="native" switchType="1"  slaveThreshold="100">
        <heartbeat>select user()</heartbeat>
      
        <writeHost host="hostM1" url=" 192.168.0.122:3306" user="root"
                   password="0123456">
            <readHost host="hostS1" url="192.168.0.125:3306" user="root" password="0123456" />
        </writeHost>
        <writeHost host="hostM2" url=" 192.168.0.142:3306" user="root"
                   password="0123456">
            <readHost host="hostS2" url="192.168.0.112:3306" user="root" password="0123456" />
        </writeHost>
    </dataHost>
</mycat:schema>

3)代码测试

操作多张数据表即可。


11,主键策略

server.xml中sequenceHandlerType值

本地0,数据库1,时间戳2

1)本地

修改值为0后,直接运行报错 mycat sequnce err.java.lang.NumberFormatException: null

修改

主键存储在

[root@localhost mycat]# cat conf/sequence_conf.properties
。。。
GLOBAL.CURID=10000

。。。。

报错:

Can not read response from server. Expected to read 4 bytes, read 0 bytes before connection was unexpectedly lost.

INSERT INTO t_order2(id,name)VALUES(next value for MYCATSEQ_GLOBAL,"李四")
> 2013 - Lost connection to server during query
> Time: 0s
直接操作mysql插入可以,通过mycat插入就不行

但是通过mysql查询可以。。。。

能读不能写,查看写入配置是不是有问题,发现IP多了空格。。。。但改完后依然同样问题。


2)数据库方式

配置文件对应的是

[root@localhost mycat]# cat conf/sequence_db_conf.properties
#sequence stored in datanode
GLOBAL=dn1
COMPANY=dn1
CUSTOMER=dn1
ORDERS=dn1
[root@localhost mycat]#


3)时间戳方式

之前测试的方式。


特么的,不测了,MyCat基本每做一步都出一个bug。。。。


--
senRsl
2022年07月02日21:06:11

没有评论 :

发表评论