CentOS 7 mongodb_3.4 sharding
环境准备
系统版本：CentOS Linux release 7.4.1708
服务器IP地址
172.16.8.24 Node1 172.16.8.25 Node2 172.16.8.26 Node3
服务器的规划如下图
MongoDB的下载安装
mongodb的版本地址：https://www.mongodb.org/dl/linux/x86_64 本次使用下载的版本 # wget http://downloads.mongodb.org/linux/mongodb-linux-x86_64-v3.4-latest.tgz # tar xf mongodb-linux-x86_64-v3.4-latest.tgz # mkdir /usr/local/mongodb # mv mongodb-linux-x86_64-3.4.17-13-g5099801229/* /usr/local/mongodb
配置环境变量
# vim /etc/profile 追加内容 export MONGODB_HOME=/usr/local/mongodb export PATH=$MONGODB_HOME/bin:$PATH 使立即生效 # source /etc/profile 三台服务器使用使用hostname访问 # vim /etc/hosts 追加内容 172.16.8.24 node1 172.16.8.25 node2 172.16.8.26 node3
创建mongodb的相关目录(conf configsvr mongos db shard1 shard2 shard3 log)
mkdir -p /data/mongodb/log mkdir -p /data/mongodb/conf mkdir -p /data/mongodb/mongos mkdir -p /data/mongodb/db/{shard1,shard2,shard3}
集群配置
1.config server配置服务器
# vim /data/mongodb/conf/configsvr.conf 配置内容 ## content systemLog: destination: file logAppend: true path: /data/mongodb/log/configsvr.log # Where and how to store data. storage: dbPath: /data/mongodb/configsvr journal: enabled: true # how the process runs processManagement: fork: true pidFilePath: /data/mongodb/configsvr/configsrv.pid # network interfaces net: port: 27002 bindIp: 0.0.0.0 #operationProfiling: replication: replSetName: config sharding: clusterRole: configsvr
启动三台服务器的config server
mongod -f /data/mongodb/conf/configsvr.conf
登录任意一台配置服务器，初始化配置副本集
# mongo node1:27002 # > use admin; config变量 # > config = {_id : "config",members : [{_id : 0, host : "node1:27002" },{_id : 1, host : "node2:27002" },{_id : 2, host : "node3:27002" }]} 初始化副本集 # > rs.initiate(config) 查看集群状态 # >rs.status();
2.配置分片副本集(三台机器)
设置第一个分片副本集
# vim /data/mongodb/conf/shard1.conf 配置内容 # where to write logging data. systemLog: destination: file logAppend: true path: /data/mongodb/log/shard1.log # Where and how to store data. storage: dbPath: /data/mongodb/db/shard1 journal: enabled: true directoryPerDB: true engine: wiredTiger wiredTiger: engineConfig: cacheSizeGB: 3 #设置使用内存大小 statisticsLogDelaySecs: 0 journalCompressor: snappy directoryForIndexes: true collectionConfig: blockCompressor: snappy indexConfig: prefixCompression: true # how the process runs processManagement: fork: true pidFilePath: /data/mongodb/db/shard1/shard1.pid # network interfaces net: port: 27010 bindIp: node1 #根据实际情况进行调整 #operationProfiling: operationProfiling: slowOpThresholdMs: 100 mode: slowOp replication: oplogSizeMB: 2048 replSetName: shard1 secondaryIndexPrefetch: all sharding: clusterRole: shardsvr
启动shard1分片
# mongod -f /data/mongodb/conf/shard1.conf
初始化shard1分片的集合
# mongo node1:27010 # > use admin; config变量 # >config = {_id : "shard1",members : [{_id : 0, host : "node1:27010",priority:5 },{_id : 1, host : "node2:27010",priority:3 },{_id : 2, host : "node3:27010",arbiterOnly: true }]} 初始化副本集 # > rs.initiate(config) 查看集群状态 # >rs.status();
设置第二个分片副本集
# vim /data/mongodb/conf/shard2.conf 配置内容 # where to write logging data. systemLog: destination: file logAppend: true path: /data/mongodb/log/shard2.log # Where and how to store data. storage: dbPath: /data/mongodb/db/shard2 journal: enabled: true directoryPerDB: true engine: wiredTiger wiredTiger: engineConfig: cacheSizeGB: 3 #根据实际情况设置内存大小 statisticsLogDelaySecs: 0 journalCompressor: snappy directoryForIndexes: true collectionConfig: blockCompressor: snappy indexConfig: prefixCompression: true # how the process runs processManagement: fork: true pidFilePath: /data/mongodb/db/shard2/shard2.pid # network interfaces net: port: 27020 bindIp: node1 #根据实际情况绑定IP #operationProfiling: operationProfiling: slowOpThresholdMs: 100 mode: slowOp replication: oplogSizeMB: 2048 replSetName: shard2 secondaryIndexPrefetch: all sharding: clusterRole: shardsvr
启动shard2分片
# mongod -f /data/mongodb/conf/shard2.conf
初始化shard2分片
# mongo node3:27020 # > use admin; config变量 # >config = {_id : "shard2",members : [{_id : 0, host : "node1:27020",priority:3 },{_id : 1, host : "node2:27020",arbiterOnly: true },{_id : 2, host : "node3:27020",priority:5 }]} 初始化副本集 # > rs.initiate(config) 查看集群状态 # >rs.status();
设置第三个分片副本集
# vim /data/mongodb/conf/shard3.conf 配置内容 # where to write logging data. systemLog: destination: file logAppend: true path: /data/mongodb/log/shard3.log # Where and how to store data. storage: dbPath: /data/mongodb/db/shard3 journal: enabled: true directoryPerDB: true engine: wiredTiger wiredTiger: engineConfig: cacheSizeGB: 3 #根据实际情况设置内存大小 statisticsLogDelaySecs: 0 journalCompressor: snappy directoryForIndexes: true collectionConfig: blockCompressor: snappy indexConfig: prefixCompression: true # how the process runs processManagement: fork: true pidFilePath: /data/mongodb/db/shard3/shard3.pid # network interfaces net: port: 27030 bindIp: node1 #根据实际情况绑定IP #operationProfiling: operationProfiling: slowOpThresholdMs: 100 mode: slowOp replication: oplogSizeMB: 2048 replSetName: shard3 secondaryIndexPrefetch: all sharding: clusterRole: shardsvr
启动shard3分片
# mongod -f /data/mongodb/conf/shard3.conf
初始化shard3分片
# mongo node2:27030 # > use admin; config变量 # >config = {_id : "shard3",members : [{_id : 0, host : "node1:27030",arbiterOnly: true },{_id : 1, host : "node2:27030",priority:5 },{_id : 2, host : "node3:27030",priority:3 }]} 初始化副本集 # > rs.initiate(config) 查看集群状态 # >rs.status();
3.配置路由服务器 mongos
# vim /data/mongodb/conf/mongos.conf 配置内容 # content systemLog: destination: file logAppend: true path: /data/mongodb/log/mongos.log processManagement: fork: true pidFilePath: /data/mongodb/mongos/mongos.pid # network interfaces net: port: 27001 bindIp: node1 #根据实际情况绑定IP sharding: configDB: config/node1:27002,node2:27002,node3:27002
启动三台服务器的mongos server
# mongos -f /data/mongodb/conf/mongos.conf
连接路由服务器和shard分片副本集
# mongo node1:27001 # > use admin #串联路由服务器与分片副本集 > sh.addShard("shard1/node1:27010,node2:27010,node3:27010") > sh.addShard("shard2/node1:27020,node2:27020,node3:27020") > sh.addShard("shard3/node1:27030,node2:27030,node3:27030") #查看集群状态 > sh.status()
测试验证
插入数据后，数据能够自动分片，指定需要激活分片的数据库和集合 数据库testdb激活分片 db.runCommand( { enablesharding :"testdb"}); 指定数据库里需要分片的集合和片键，哈希id 分片 db.runCommand( { shardcollection : "testdb.table1",key : {"id": "hashed"} } );
测试分片配置是否正确
# mongo node1:27001 切换数据库到testdb >use testdb 插入50万数据 >for(i=1;i<=500000;i++){db.table1.insert({"id":i,"name":"linux"})}; 查看分片信息 > db.table1.stats();
模拟三台服务器其中一台故障
将服务器node3 关机 # shutdown -h now 检查集群状态 #sh.status() 继续写入50万数据到testdb数据库 > use testdb 插入50万数据 >for(i=1;i<=500000;i++){db.table1.insert({"id":i,"name":"mongodb"})}; 查询表数据条数应该有100万 > db.table1.count()
mongodb数据库添加安全权限访问控制
进入mongodb的mongos # mongo node3:27001 > use admin 创建超级管理员 > db.createUser({user:"root",pwd:"centosmongodb",roles:[{role:"root",db:"admin"}]}) 验证 > db.auth("root","centosmongodb") 在testdb创建一个读写权限的账号 > use testdb; > db.createUser({user:"test",pwd:"testdb",roles:[{role:"readWrite",db:"testdb"}]}) > db.auth("test","testdb")
生成keyfile.key，用于集群的通信验证，拷贝文件到其他几台服务器
# cd /data/mongodb/conf # openssl rand -base64 756 >keyfile.key # chmod 400 keyfile.key
在所有服务器上面停止mongodb的进程
# killall mongos # killall mongod
括号里面的配置文件追加认证的的配置（configsvr.conf shard1.conf shard2.conf shard3.conf）
security: keyFile: /data/mongodb/conf/keyfile.key clusterAuthMode: keyFile authorization: enabled
mongos 配置文件有些特殊，只需要添加下面的配置（mongos.conf）
security: keyFile: /data/mongodb/conf/keyfile.key clusterAuthMode: keyFile
启动mongodb的进程，先启动配置服务器，然后启动分片，最后启动mongos
# mongod -f /data/mongodb/conf/configsvr.conf # mongod -f /data/mongodb/conf/shard1.conf # mongod -f /data/mongodb/conf/shard2.conf # mongod -f /data/mongodb/conf/shard3.conf # mongos -f /data/mongodb/conf/mongos.conf
浙公网安备 33010602011771号