mongodb实现一主两从一个仲裁者

4台机器安装mongodb后，把4台机器的mongod启动，然后通过mongo连接你要做主机的mongo。

如以192.168.10.111:27017为主节点

master.conf配置

processManagement:

   fork: true

net:

   bindIp: 127.0.0.1,192.168.10.110,10.9.0.1,14.17.97.226

   port: 27017

storage:

   dbPath: /mnt/raid5/server/data/mongodb

systemLog:

   destination: file

   path: "/mnt/raid5/server/logs/mongodb/mongodb.log"

   logAppend: true

   logRotate: rename

storage:

   journal:

      enabled: true

replication:

   replSetName: rs01

其他主机上mongo配置一样只需修改一下bindIP

#mongo 192.168.10.111:27017

然后修改mongodb配置

>use admin

>cfg={ _id:"rs01", members:[ {_id:0,host:'192.168.10.111:27017',priority:5}, {_id:1,host:'192.168.10.110:29017',priority:3},{_id:2,host:'192.168.10.112:27017',priority:2},{_id:3,host:'192.168.10.114:27017',arbiterOnly:true}] }

>rs.reconfig(cfg)

错误

> rs.reconfig(cfg)

{

        "ok" : 0,

        "errmsg" : "replSetReconfig should only be run on PRIMARY, but my state is REMOVED; use the \"force\" argument to override",

        "code" : 10107

}

解决

> rs.reconfig(cfg,{"force":true})

错误

> rs.reconfig(cfg)

2016-04-23T15:41:01.557+0800 E QUERY    Error: Could not retrieve replica set config: {

        "info" : "run rs.initiate(...) if not yet done for the set",

        "ok" : 0,

        "errmsg" : "no replset config has been received",

        "code" : 94

}

    at Function.rs.conf (src/mongo/shell/utils.js:1017:11)

    at Function.rs.reconfig (src/mongo/shell/utils.js:969:22)

    at (shell):1:4 at src/mongo/shell/utils.js:1017

解决

rs.initiate(cfg)

数值越大优先级越高，即priority值最大的为主节点

设置优先级

>use admin

>config=rs.conf()

>config.members[0].priority=5

>rs.reconfig(config,{force:true})

关闭mongod

#./mongo

>use admin

>db.shutdownServer()

启动mongodb脚本：
start_mongo_master.sh
#!/bin/sh
su mongodb -c "numactl --interleave=all /server/env/mongodb/bin/mongod --config /server/env/mongodb/mongod.conf --httpinterface --rest"



设置mongo打开文件数和打开程序书限制
/etc/pam.d/login文件中添加以下一行：
session    required     /lib64/security/pam_limits.so    //添加了这个，在limits.conf中添加限制用户资源马上生效

/etc/security/limits.conf文件添加以下限制
mongodb         soft    nofile          65535
mongodb         hard    nofile          65535
mongodb         soft    nproc           65536
mongodb         hard    nproc           65536

mongodb告警

2016-05-05T15:24:01.461+0800 I STORAGE  [initandlisten] 

2016-05-05T15:24:01.461+0800 I STORAGE  [initandlisten] ** WARNING: Readahead for /mnt/raid5/server/data/mongodb_master is set to 2048KB

2016-05-05T15:24:01.461+0800 I STORAGE  [initandlisten] **          We suggest setting it to 256KB (512 sectors) or less

2016-05-05T15:24:01.461+0800 I STORAGE  [initandlisten] **          http://dochub.mongodb.org/core/readahead

解决：

RO    RA   SSZ   BSZ   StartSec            Size   Device

rw   256   512  4096          0   1000204886016   /dev/sda

rw   256   512  4096       2048      2097152000   /dev/sda1

rw   256   512  4096    4098048     33554432000   /dev/sda2

rw   256   512  4096   69634048       209715200   /dev/sda3

rw   256   512  4096   70043648    964339925504   /dev/sda4

rw   256   512  4096          0   1000204886016   /dev/sdc

rw   256   512   512         63    500096991744   /dev/sdc1

rw   256   512  4096  976752000    500097024000   /dev/sdc2

rw   256   512  4096          0   1000204886016   /dev/sdb

rw   256   512   512         63    500096991744   /dev/sdb1

rw   256   512  4096  976752000    500097024000   /dev/sdb2

rw   2048   512  4096          0    999925219328   /dev/md0

把/dev/md0的readahead设置成256就行了

#blockdev --setra 256 /dev/md0

记得在rc.local中添加blockdev --setra 256 /dev/md0，要不然他每次都会恢复默认值




mongodb告警
** WARNING: /sys/kernel/mm/transparent_hugepage/enabled is 'always'.
 **        We suggest setting it to 'never'

** WARNING: /sys/kernel/mm/transparent_hugepage/defrag is 'always'.
**        We suggest setting it to 'never'
解决：
/etc/rc.local中添加
echo "never" > /sys/kernel/mm/transparent_hugepage/enabled
echo "never" > /sys/kernel/mm/transparent_hugepage/defrag
这两行

自动备份数据库脚本

#!/bin/sh

host=192.168.10.110

port=27017

cmdexec="/server/env/mongodb/bin/mongodump" 

db="Bigdata"

outdbpath="/server/backup/wechat_backup/"

Date=`date +%Y%m%d%H`

$cmdexec -h $host --port $port -d $db -o $outdbpath$db

tar zcvf $outdbpath$db$Date.tar.gz  $db

rm -rf  $db

自动清理最近三天备份数据脚本

#!/bin/sh

## clean for 3 days before ,which file have tar.gz

find ./ -name "*.tar.gz" -mtime +2 -exec rm -r {} \;

只有一个主机做mongo

#mongo --port 29017

>use admin

>cfg={_id:"rs1",members:[{_id:0,host:'192.168.0.212:29017',priority:1}]}

>rs.initiate(cfg)


现在mongo内存使用，mongodb3.2版本以上才有
storage:
    journal:
        enabled: true
        commitIntervalMs: 500
    directoryPerDB: true
    dbPath: "/server/env/mongodb-linux-x86_64-3.2.8/database"    
    engine: wiredTiger    
    wiredTiger:
        engineConfig:
            directoryForIndexes: true
            cacheSizeGB: 10
        collectionConfig:
            configString: "type=lsm"
        indexConfig:
            configString: "type=lsm"