搭建MongoDB复制集

发表于 分类于 本文字数： 6.1k 阅读时长 ≈ 6 分钟

目前的MongoDB还是单机使用，虽然有用mongoexport导出数据备份，用着还是慌…学着搭个复制集看看吧

简述复制集

MongoDB复制集的意义在于实现服务高可用、数据分发、读写分离、异地容灾，依赖于两个功能实现:

  • 数据写入时将数据迅速复制到另一个独立节点上
  • 在接受写入的节点发生故障时自动选举出一个新的替代节点(复制集由3个以上具有投票权的节点组成，一般为奇数节点)

准备节点

1、官方手册中有命令行安装的教程:
https://docs.mongodb.com/manual/tutorial/install-mongodb-on-ubuntu/#overview
由于我在机器中已经按步骤安装过一个MongoDB，又是在单机学习搭建复制的步骤，就采用创建临时目录，用MongoDB压缩包的形式测试下: 

1
2
3
4
5
$ mkdir MongoWork/MongoData/db{1,2,3}
$ cd MongoWork
$ wget https://fastdl.mongodb.org/linux/mongodb-linux-x86_64-ubuntu1804-4.2.1.tgz
$ tar -xvf mongodb-linux-x86_64-ubuntu1804-4.2.1.tgz
$ mv mongodb-linux-x86_64-ubuntu1804-4.2.1 mongodb-4.2.1

在三个db文件夹下分别创建mongo.conf文件，并写入内容: 

1
2
3
4
5
6
7
8
9
10
11
12
13
systemLog:
  destination: file
  path: /home/top/MongoWork/MongoData/db1/mongod.log
  logAppend: true
storage:
  dbPath: /home/top/MongoWork/MongoData/db1
net:
  bindIp: 0.0.0.0
  port: 28018
replication:
  replSetName: rs0
processManagement:
  fork: true

注意上面的path、dbPath、port要分别修改。

现在分别启动三个mongodb: 

1
2
3
4
5
6
7
8
9
10
11
12
13
$ cd mongodb-4.2.1
$ bin/mongod -f ../MongoData/db1/mongod.conf 
about to fork child process, waiting until server is ready for connections.
forked process: 3714
child process started successfully, parent exiting
$ bin/mongod -f ../MongoData/db2/mongod.conf 
about to fork child process, waiting until server is ready for connections.
forked process: 3752
child process started successfully, parent exiting
$ bin/mongod -f ../MongoData/db3/mongod.conf 
about to fork child process, waiting until server is ready for connections.
forked process: 3790
child process started successfully, parent exiting

通过ps和netstat分别查看一下: 

1
2
3
4
5
6
7
8
9
10
11
12
$ ps -ef | grep mongod
top        3714      1  3 22:09 ?        00:00:00 bin/mongod -f ../MongoData/db1/mongod.conf
top        3752      1  6 22:09 ?        00:00:01 bin/mongod -f ../MongoData/db2/mongod.conf
top        3790      1 16 22:09 ?        00:00:01 bin/mongod -f ../MongoData/db3/mongod.conf
top        3827   3111  0 22:09 pts/0    00:00:00 grep --color=auto mongod

$ netstat -antp | grep mongod
(Not all processes could be identified, non-owned process info
 will not be shown, you would have to be root to see it all.)
tcp        0      0 0.0.0.0:28018           0.0.0.0:*               LISTEN      3714/bin/mongod     
tcp        0      0 0.0.0.0:28019           0.0.0.0:*               LISTEN      3752/bin/mongod     
tcp        0      0 0.0.0.0:28020           0.0.0.0:*               LISTEN      3790/bin/mongod

配置节点

进入第一个节点: 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
$ bin/mongo top:28018
> rs.initiate()
{
    "info2" : "no configuration specified. Using a default configuration for the set",
    "me" : "top:28018",
    "ok" : 1,
    "$clusterTime" : {
        "clusterTime" : Timestamp(1575211318, 1),
        "signature" : {
            "hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
            "keyId" : NumberLong(0)
        }
    },
    "operationTime" : Timestamp(1575211318, 1)
}
rs0:SECONDARY> 
rs0:PRIMARY>

信息中显示的rs0:SECONDARY已经进入复制集的从节点状态，rs0:PRIMARY显示成为了主节点。这个时候查看节点的状态: 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
rs0:PRIMARY> rs.status()
{
    "set" : "rs0",
    "date" : ISODate("2019-12-01T14:44:00.477Z"),
    "myState" : 1,
    "term" : NumberLong(1),
    "syncingTo" : "",
    "syncSourceHost" : "",
    "syncSourceId" : -1,
    "heartbeatIntervalMillis" : NumberLong(2000),
    "majorityVoteCount" : 1,
    "writeMajorityCount" : 1,
    ......
    "members" : [
        {
            "_id" : 0,
            "name" : "top:28018",
            "ip" : "127.0.1.1",
            "health" : 1,
            "state" : 1,
            "stateStr" : "PRIMARY",
            "uptime" : 163,
            "optime" : {
                "ts" : Timestamp(1575211439, 1),
                "t" : NumberLong(1)
            },
            "optimeDate" : ISODate("2019-12-01T14:43:59Z"),
            "syncingTo" : "",
            "syncSourceHost" : "",
            "syncSourceId" : -1,
            "infoMessage" : "",
            "electionTime" : Timestamp(1575211318, 2),
            "electionDate" : ISODate("2019-12-01T14:41:58Z"),
            "configVersion" : 1,
            "self" : true,
            "lastHeartbeatMessage" : ""
        }
    ],
    "ok" : 1,
    .......

members表明了复制集中的成员信息，此时还是一个单节点，接下来加入其它两个节点: 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
rs0:PRIMARY> rs.add("top:28019")
{
    "ok" : 1,
    "$clusterTime" : {
        "clusterTime" : Timestamp(1575211629, 1),
        "signature" : {
            "hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
            "keyId" : NumberLong(0)
        }
    },
    "operationTime" : Timestamp(1575211629, 1)
}
rs0:PRIMARY> rs.add("top:28020")
{
    "ok" : 1,
    "$clusterTime" : {
        "clusterTime" : Timestamp(1575211634, 1),
        "signature" : {
            "hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
            "keyId" : NumberLong(0)
        }
    },
    "operationTime" : Timestamp(1575211634, 1)
}
rs0:PRIMARY>

再次键入rs.status()可以看到members中已经三个节点信息了。

验证复制集

在主节点中写入信息: 

1
2
3
4
rs0:PRIMARY> db.test.insert({ a: 1})
WriteResult({ "nInserted" : 1 })
rs0:PRIMARY> db.test.insert({ b: 2})
WriteResult({ "nInserted" : 1 })

在另外两个会话窗口中分别键入命令进入mongodb: 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
$ bin/mongo top:28019
rs0:SECONDARY> db.test.find()
Error: error: {
    "operationTime" : Timestamp(1575211979, 1),
    "ok" : 0,
    "errmsg" : "not master and slaveOk=false",
    "code" : 13435,
    "codeName" : "NotMasterNoSlaveOk",
    "$clusterTime" : {
        "clusterTime" : Timestamp(1575211979, 1),
        "signature" : {
            "hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
            "keyId" : NumberLong(0)
        }
    }
}
rs0:SECONDARY>

显示报错，不允许在从节点读，此时键入rs.slaveOk()即可: 

1
2
3
4
5
rs0:SECONDARY> rs.slaveOk()
rs0:SECONDARY> db.test.find()
{ "_id" : ObjectId("5de3d338198c429c8f8587c0"), "a" : 1 }
{ "_id" : ObjectId("5de3d343198c429c8f8587c1"), "b" : 2 }
rs0:SECONDARY>

可以看到数据已经进行了复制。

关于NTP

使用复制集的话，需要注意在各个节点配置下NTP，如果运行ntpd服务，一般来说ntpd会逐渐调整时钟，避免时间跳变。没有的话可以先安装，然后在ntp.conf中增加一行:

1
server ntp7.aliyun.com

如果你的机器的时钟发生跳变不会有严重后果，可以通过ntpdate ntp7.aliyun.com进行一次性的同步。

参考

1、https://docs.mongodb.com/manual/administration/production-notes/#recommended-configuration
2、https://tuna.moe/help/ntp/
3、https://help.aliyun.com/document_detail/92704.html