MongoDB highly available deployment

规划部署情况如下:

1、准备基础环境

在网上下载mongodb-linux-x86_64-rhel62-3.4.2.gz放到各个服务器上，解压后放在对应的目录下，然后在根目录下新建data文件夹统一用于存放数据和配置信息。

2、搭建配置集群243-244-245

在mongo的根目录下新建start脚本，内容如下

bin/./mongod --dbpath data/ --logpath data/mongo.log --configsvr --fork --port 20243 --replSet config

其中 configsvr表示配置服务，fork表示后台启动，replSet表示集群，config表示集群名字

同样在其他配置服务上也建立对应的start启动文件，记得修改端口哦。

启动-3台都启动

./start

随便进入一台机器，进入mongo，运行如下命令

>config = { _id:"config", configsvr:true, members:[ {_id:0,host:'172.16.13.243:20243'}, {_id:1,host:'172.16.13.244:20244'}, {_id:2,host:'172.16.13.245:20245'} ] }{"_id" : "config","configsvr" : true,"members" : [{"_id" : 0,"host" : "172.16.13.243:20243"},{"_id" : 1,"host" : "172.16.13.244:20244"},{"_id" : 2,"host" : "172.16.13.245:20245"}]}> rs.initiate(config){ "ok" : 1 }

查看集群状态

config:PRIMARY> rs.status();{"set" : "config","date" : ISODate("2017-04-12T09:29:51.889Z"),"myState" : 1,"term" : NumberLong(1),"configsvr" : true,"heartbeatIntervalMillis" : NumberLong(2000),"optimes" : {"lastCommittedOpTime" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"readConcernMajorityOpTime" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"appliedOpTime" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"durableOpTime" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)}},"members" : [{"_id" : 0,"name" : "172.16.13.243:20243","health" : 1,"state" : 1,"stateStr" : "PRIMARY","uptime" : 621,"optime" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"optimeDate" : ISODate("2017-04-12T09:29:42Z"),"infoMessage" : "could not find member to sync from","electionTime" : Timestamp(1491989360, 1),"electionDate" : ISODate("2017-04-12T09:29:20Z"),"configVersion" : 1,"self" : true},{"_id" : 1,"name" : "172.16.13.244:20244","health" : 1,"state" : 2,"stateStr" : "SECONDARY","uptime" : 42,"optime" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"optimeDurable" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"optimeDate" : ISODate("2017-04-12T09:29:42Z"),"optimeDurableDate" : ISODate("2017-04-12T09:29:42Z"),"lastHeartbeat" : ISODate("2017-04-12T09:29:50.238Z"),"lastHeartbeatRecv" : ISODate("2017-04-12T09:29:51.467Z"),"pingMs" : NumberLong(0),"syncingTo" : "172.16.13.243:20243","configVersion" : 1},{"_id" : 2,"name" : "172.16.13.245:20245","health" : 1,"state" : 2,"stateStr" : "SECONDARY","uptime" : 42,"optime" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"optimeDurable" : {"ts" : Timestamp(1491989382, 1),"t" : NumberLong(1)},"optimeDate" : ISODate("2017-04-12T09:29:42Z"),"optimeDurableDate" : ISODate("2017-04-12T09:29:42Z"),"lastHeartbeat" : ISODate("2017-04-12T09:29:50.237Z"),"lastHeartbeatRecv" : ISODate("2017-04-12T09:29:51.537Z"),"pingMs" : NumberLong(0),"syncingTo" : "172.16.13.243:20243","configVersion" : 1}],"ok" : 1}

3、搭建分片集群246-247

在246机器的mongo的根目录下新建start脚本，内容如下

[root@localhost mongodb-32246]# cat start bin/./mongod --dbpath data/ --logpath data/mongo.log --fork --port 32246 --replSet 246 --shardsvr

replSet代表集群，shardsvr代表分片，246代表集群的名字，在246的其他mongo目录下也建立对应的start脚本，记得修改端口，启动脚本。

运行如下命令

config = {... _id:'246',... members:[... {_id:0,host:'172.16.13.246:30246'},... {_id:1,host:'172.16.13.246:31246'},... {_id:2,host:'172.16.13.246:32246'}... ]... }{"_id" : "246","members" : [{"_id" : 0,"host" : "172.16.13.246:30246"},{"_id" : 1,"host" : "172.16.13.246:31246"},{"_id" : 2,"host" : "172.16.13.246:32246"}]}> rs.initiate(config){ "ok" : 1 }

查看集群状态

这里不在列出，同样的操作在247上执行一次，记得修改集群名字

4、搭建mongos

在241的mongos根目录下建立start脚本

[root@localhost mongodb-3.4.2]# cat start bin/./mongos --logpath data/mongo.log --fork --configdb config/172.16.13.243:20243,172.16.13.244:20244,172.16.13.245:20245

进入mongos的命令界面，

添加分片

mongos> sh.addShard("246/172.16.13.246:30246,172.16.13.246:31246,172.16.13.246:32246"){ "shardAdded" : "246", "ok" : 1 }mongos> sh.addShard("247/172.16.13.247:30247,172.16.13.247:31247,172.16.13.247:32247"){ "shardAdded" : "247", "ok" : 1 }

在242上也运行上述命令。

设置test.aj表根据_id字段按照hash方式分配数据

mongos> sh.enableSharding("test"){ "ok" : 1 }mongos> sh.shardCollection("test.aj", { _id: 'hashed'}){ "collectionsharded" : "test.aj", "ok" : 1 }

5、插入数据

for (var id = 1; id db.aj.count();5082

246子节点验证数据量

246:SECONDARY> use testswitched to db test246:SECONDARY> db.getMongo().setSlaveOk();246:SECONDARY> db.aj.count();5082

247重复上述操作

247:PRIMARY> db.aj.count();4918247:SECONDARY> use test;switched to db test247:SECONDARY> db.getMongo().setSlaveOk();247:SECONDARY> db.aj.count();4918