Orchestrator部署

数据运维 2024-03-26 大白菜程序猿手机阅读

Orchestrator部署

orc元数据使用的数据库 + raft + orc软件

192.168.0.216
192.168.0.217
192.168.0.218

被监控mysql

192.168.0.211
192.168.0.212
192.168.0.213

下载

https://github.com/github/orchestrator/releases

具体选择v3.2.6

#服务包wget wget https://github.com/openark/orchestrator/releases/download/v3.2.6/orchestrator-3.2.6-1.x86_64.rpm #二进制命令wget wget https://github.com/openark/orchestrator/releases/download/v3.2.6/orchestrator-cli-3.2.6-1.x86_64.rpm #命令行工具 wget https://github.com/openark/orchestrator/releases/download/v3.2.6/orchestrator-client-3.2.6-1.x86_64.rpm

文件校验

检测是否下载完毕，vm网络环境不好，第一次安装的时候报错居然是没下载完整 sha256sum orchestrator-client-3.2.6-1.x86_64.rpm sha256sum orchestrator-3.2.6-1.x86_64.rpm sha256sum orchestrator-cli-3.2.6-1.x86_64.rpm

安装

rpm -ivh orchestrator-3.2.6-1.x86_64.rpm rpm -ivh orchestrator-client-3.2.6-1.x86_64.rpm rpm -ivh orchestrator-cli-3.2.6-1.x86_64.rpm

sqlite存储元数据

注:配置文件如下添加即可

vi /usr/local/orchestrator/orchestrator.conf.json "BackendDB": "sqlite", "SQLite3DataFile": "/usr/local/orchestrator/orchestrator.sqlite3",

mysql存储元数据（本次暂时不用，VM安装用sqllite）

CREATE DATABASE IF NOT EXISTS orchestrator; CREATE USER 'orch_manager'@'%' IDENTIFIED BY '123456'; GRANT ALL ON orchestrator.* TO 'orch_manager'@'%'; GRANT SUPER, REPLICATION SLAVE, REPLICATION CLIENT ON *.* TO 'orch_manager'@'%';

被监控的mysql上执行

set sql_log_bin=0; -- drop user 'orc_client_user'@'%' ; CREATE USER 'orc_client_user'@'%' IDENTIFIED with mysql_native_password by 'orc_client_password'; GRANT SUPER, PROCESS, REPLICATION SLAVE, REPLICATION CLIENT, RELOAD ON *.* TO 'orc_client_user'@'%'; set sql_log_bin=1;

为避免V8.0登录验证报错，使用mysql_native_password 方式验证

CREATE USER 'orc_client_user'@'%' IDENTIFIED BY 'orc_client_passwor'; GRANT SUPER, PROCESS, REPLICATION SLAVE, REPLICATION CLIENT, RELOAD ON *.* TO 'orc_client_user'@'%';


CREATE DATABASE meta;

GRANT SELECT ON meta.* TO 'orc_client_user'@'%';

### 配置启动

cd /usr/local/orchestrator/ #mysql cp orchestrator-sample.conf.json orchestrator.conf.json #sqllite cp orchestrator-sample-sqlite.conf.json orchestrator.conf.json

编辑

#vi /usr/local/orchestrator/orchestrator.conf.json

RaftBind MySQLOrchestratorHost 修改成各自节点的值

MySQLTopologyUser 修改实际用户

#mysql配置 "MySQLTopologyUser": "orc_client_user", "MySQLTopologyPassword": "orc_client_password", #后端mysql "MySQLOrchestratorHost": "192.168.0.216", "MySQLOrchestratorPort": 3306, "MySQLOrchestratorDatabase": "orchestrator", "MySQLOrchestratorUser": "orchestrator", "MySQLOrchestratorPassword": "mysql",

"RaftEnabled": true, "RaftDataDir": "/data/orchestrator", "RaftBind": "192.168.0.216", "DefaultRaftPort": 10008, "RaftNodes": [ "192.168.0.216", "192.168.0.217", "192.168.0.218" ],

以下2项需要设置为"*",否则不会发生启动切换；另 /data/orchestrator 让orc自动创建

"RecoverMasterClusterFilters": [ "*" ], "RecoverIntermediateMasterClusterFilters": [ "*" ],

用户密码登录

"HTTPAuthUser": "admin", "HTTPAuthPassword": "Jsxfws@12",

启动：

cd /usr/local/orchestrator ./orchestrator -config ./orchestrator.conf.json http >> ./orchestrator.log 2>&1 &

查看状态

http://192.168.0.218:3000/web/status

日常操作

注意：

gtid模式下，change master 需设置Auto_Position=1

set global slave_net_timeout=8; 约8~10秒感知主库异常
change master MASTER_CONNECT_RETRY=1 主库1秒无数据变更，备库会尝试重连主库

web discover

Orchestrator部署-每日运维

orchestrator-client

# 列出集群，集群别名 orchestrator-client -c api -path clusters orchestrator-client -c clusters-alias 显示别名为vm-test1的集群的master # 列出别名为"node02"的集群 orchestrator-client -c which-cluster-master -alias node02 #显示别名为node02的集群下所有的实例 orchestrator-client -c which-cluster-instances -alias node02

web方式遇到报错无反馈，使用命令行方式discovery：

orchestrator-client -c discover -i 192.168.0.211:3306

注：orc会通过dns lookup ip会受影响 (如在VM中设置dns1=114.114.114.114)

故障检测

自动故障切换

在主库上执行shutdown模拟主库故障

shutdown;

Orchestrator部署-每日运维

注：主库宕机，自动Failover到node02

优雅的主从切换

# orchestrator-client -c graceful-master-takeover -i node01:3306 #master由node02 切换到 node01 #orchestrator-client -c graceful-master-takeover --alias node02 -d node01:3306

手动强制切换

#orchestrator-client -c force-master-failover --alias node01 -i node02:3306

Orchestrator部署-每日运维

原主被分离，可以手动处理后再加入