orchestrator简介

谁践踏了优雅 2022-06-10 08:42 149阅读 0赞

github：  
[https://github.com/github/orchestrator/tree/master/docs][https_github.com_github_orchestrator_tree_master_docs]

Orchestrator介绍  
[https://www.percona.com/blog/2016/03/08/orchestrator-mysql-replication-topology-manager/][https_www.percona.com_blog_2016_03_08_orchestrator-mysql-replication-topology-manager]

mysql多实例    
    1、初始化mysql到新的数据目录
        cd /usr/local/mysql-5.5.45;./scripts/mysql_install_db --user=mysql --datadir=/var/lib/mysql_3307
    2、复制当前配置文件，修改下列参数
        cp  /etc/my.cnf  /etc/my2.cnf
        vim /etc/my2.cnf
            [mysqld]
            datadir=/var/lib/mysql_3307                                  //数据存放目录
    3、复制/etc/init.d/mysqld
            cp /etc/init.d/mysqld /etc/init.d/mysqld_3307
            修改 /etc/init.d/mysqld_3307 中默认配置文件位置 Defaults_file="/etc/my2.cnf"

（1）mysql主从  
主

grant replication slave on *.* to 'vera'@'192.168.3.146' identified by '123456';
        show master status;
        从
        stop slave;
        change master to  master_host='192.168.3.87',master_port=3307,master_user='vera', master_password='123456', master_log_file='mysql-bin.000008', master_log_pos=107;
        start slave;
        show slave status\G;
    
        change master to  master_host='192.168.3.146',master_port=3307,master_user='vera', master_password='123456', master_log_file='mysql-bin.000010', master_log_pos=107;

注意开放防火墙，否则连接不到3306端口

-A INPUT -m state --state NEW -m tcp -p tcp --dport 3306 -j ACCEPT
    
        drop user root@'127.0.0.1';

（2）设置域名  
修改主机名

hostname rep1 
        vim /etc/sysconfig/network

设置hosts

vim /etc/hosts
                192.168.3.146 rep2
                192.168.3.87 rep3

（3）安装orchestrator服务  
1、

tar xf orchestrator-2.1.5.tar.gz -C /usr/local/orchestrator/
         cd /usr/local/orchestrator/

2、orchestrator将拓扑状态保存到后端数据库

CREATE DATABASE IF NOT EXISTS orchestrator;
         GRANT ALL PRIVILEGES ON `orchestrator`.* TO 'orchestrator'@'127.0.0.1' IDENTIFIED BY 'to8to123';

3、授权orchestrator连接Mysql实例

GRANT SUPER, PROCESS, REPLICATION SLAVE, RELOAD ON *.* TO 'orchestrator'@'192.168.5.203' IDENTIFIED BY 'to8to123';

4、修改配置文件

configure MYSQL:
        #cat /etc/orchestrator.conf.json 
        "MySQLTopologyUser": "orchestrator",
        "MySQLTopologyPassword": "to8to123",
    
        "MySQLOrchestratorHost": "orchestrator.backend.master.com",
        "MySQLOrchestratorPort": 3306,
        "MySQLOrchestratorDatabase": "orchestrator",
        "MySQLOrchestratorUser": "orchestrator_srv",
        "MySQLOrchestratorPassword": "orc_server_password",

（4）基本命令  
c\[command\] i\[instance\]

查看集群
        $ orchestrator -c clusters
         topology1.master.instance.com:3306   
    
        $ ./orchestrator -c discover -i vera:3306 
    
        查看Mysql拓扑   
        $./orchestrator -c topology -i 127.0.0.1:22987
        $ ./orchestrator -c topology -i vera:3306 --config=/etc/orchestrator.conf.json
        vera:3306         [0s,ok,5.5.45-log,rw,ROW]
        + vera.jiang:3306 [0s,ok,5.5.45-log,rw,ROW]
    
        $ /usr/local/orchestrator/orchestrator -c forget  -i 192.168.3.146:3308
    
        重新分配拓扑
        $./orchestrator -c relocate -i 127.0.0.1:22988 -d 127.0.0.1:22987
    
        恢复DeadMaster场景
        $./orchestrator -c recover -i dead.instance.com:3306 --debug
    
        手动强制失败
        $./orchestrator -c force-master-failover -i instance.in.that.cluster
    
        分析主从复制情况
        $/usr/local/orchestrator/orchestrator -c replication-analysis
    
        $/usr/local/orchestrator/orchestrator -c clusters 
    
        /usr/local/orchestrator/orchestrator -c topology -i 192.168.3.87:3307 |tr '-' ' '|awk -F, '{print $1,$2}' |awk '{print $1,$NF}'

（5）ssh 和sudo

设置sudo
        vim /etc/sudoers.d/orch
            Defaults !requiretty
            orchuser ALL=(ALL) NOPASSWD: /usr/sbin/arping,/sbin/ip,/bin/ping
        chmod 0440 /etc/sudoers.d/orch
    
        设置ssh
        useradd -m orchuser -s /bin/bash
        ssh登录分析：
          ssh XXX@IP
          1、XXX向服务器发送一个请求
          2、服务器找到authorized_keys中对应XXX的用户和IP，服务器生成一个随机数，利用XXX发送的公钥加密后发送给XXX
          3、XXX收到加密后的随机数，找到~/.ssh/下的私钥解密，讲解密的随机数返回给服务器
          4、服务器验证随机数是否匹配，匹配则XXX登录成功

（6）  
子网卡： [http://blog.163.com/fan\_yishan/blog/static/4769221320141027635931/][http_blog.163.com_fan_yishan_blog_static_4769221320141027635931]  
关闭图像界面管理  
ip addr add 192.168.3.71/24 dev eth2 label eth2:0

错误：  
Error from server: error dialing backend: dial tcp: lookup node3 on 114.114.114.114:53: no such host  
域名解析问题，配置/etc/hosts解决

2017-08-10 09:03:06 ERROR ReadTopologyInstance(vera:3306) show slave hosts: ReadTopologyInstance(vera:3306) 'show slave hosts' returned row with <host,port>: <,3306>
    Mysql主从复制，在master上查看从的信息：show slave hosts
    http://dinglin.iteye.com/blog/1255160 
    默认不记录从的host信息，只记录端口和Server_id信息，需要设置my.cnf 中的[mysqld] report-host=IP
    
    
    2017-08-10 09:17:38 ERROR dial tcp 192.168.3.103:3306: i/o timeout
    2017-08-10 09:17:38 ERROR ReadTopologyInstance(192.168.3.103:3306): dial tcp 192.168.3.103:3306: i/o timeout
    WARNING: NamedStopwatch.Stop("instance") IsRunning is false
    网络丢包造成
    
    
    Last_IO_Error: Fatal error: The slave I/O thread stops because master and slave have equal MySQL server ids; these ids must be different for replication to work (or the --replicate-same-server-id option must be used on slave but this does not always make sense; please check the manual before using it).
    
    
    
    Error_code: 1062; handler error HA_ERR_FOUND_DUPP_KEY; the event's master log mysql-bin.000009, end_log_pos 7482377
    [1062]Mysql中主键冲突问题
    跳过错误
    stop slave;
    set global sql_slave_skip_counter=1;
    start slave;
    
    
    
    [root@rep3 orchestrator]# ./orchestrator -c relocate -i 192.168.3.146:3307 -d 192.168.3.146:3306
    2017-08-11 20:40:53 ERROR 192.168.3.146:3307 cannot replicate from 192.168.3.146:3306. Reason: instance does not have log_slave_updates enabled: 192.168.3.146:3306
    2017-08-11 20:40:53 FATAL 2017-08-11 20:40:53 ERROR 192.168.3.146:3307 cannot replicate from 192.168.3.146:3306. Reason: instance does not have log_slave_updates enabled: 192.168.3.146:3306
    设置my.cnf的 [mysqld] log_slave_updates=1
    
    
    
    [root@rep3 orchestrator]# ./orchestrator -c relocate -i 192.168.3.146:3307 -d 192.168.3.146:3306
    2017-08-11 20:44:50 ERROR Relocating 192.168.3.146:3307 below 192.168.3.146:3306 turns to be too complex; please do it manually
    2017-08-11 20:44:50 FATAL 2017-08-11 20:44:50 ERROR Relocating 192.168.3.146:3307 below 192.168.3.146:3306 turns to be too complex; please do it manually
    orchestrator重新可视化拓扑要求是一个拓扑下面的slave替换，而不是两个cluster的拓扑变化

wget 文件名过长的问题  
wget XXXX -O rename.tar.gz

Linux下静态路由修改命令  
方法一：  
添加路由  
route add -net 192.168.0.0/24 gw 192.168.0.1  
route add -host 192.168.1.1 dev 192.168.0.1  
删除路由  
route del -net 192.168.0.0/24 gw 192.168.0.1

add 增加路由  
del 删除路由  
\-net 设置到某个网段的路由  
\-host 设置到某台主机的路由  
gw 出口网关 IP地址  
dev 出口网关 物理设备名

增加默认路由  
route add default gw 192.168.0.1  
默认路由一条就够了  
route -n 查看路由表

[https_github.com_github_orchestrator_tree_master_docs]: https://github.com/github/orchestrator/tree/master/docs
[https_www.percona.com_blog_2016_03_08_orchestrator-mysql-replication-topology-manager]: https://www.percona.com/blog/2016/03/08/orchestrator-mysql-replication-topology-manager/
[http_blog.163.com_fan_yishan_blog_static_4769221320141027635931]: http://blog.163.com/fan_yishan/blog/static/4769221320141027635931/