mongodb复制集+分片生产环境实践
三台机器操作系统环境如下:
[mongodb@node1 ~]$ cat /etc/issueRed Hat Enterprise Linux Server release 6.6 (Santiago)Kernel \r on an \m[mongodb@node1 ~]$ uname -r2.6.32-504.el6.x86_64[mongodb@node1 ~]$ uname -mx86_64
架构如下图:
文字描述: 1
192.168.42.41、shard1:10001、shard2:10002、shard3:10003、configsvr:10004、mongos:10005注:shard1主节点,shard2仲裁,shard3副本192.168.42.42、shard1:10001、shard2:10002、shard3:10003、configsvr:10004、mongos:10005注:shard1副本,shard2主节点,shard3仲裁192.168.42.43、shard1:10001、shard2:10002、shard3:10003、configsvr:10004、mongos:10005注:shard1仲裁,shard2副本,shard3主节点node1:192.168.42.41node2:192.168.42.42node3:192.168.42.43
创建mongodb用户
[root@node1 ~]# groupadd mongodb[root@node1 ~]# useradd -g mongodb mongodb[root@node1 ~]# mkdir /data[root@node1 ~]# chown mongodb.mongodb /data -R[root@node1 ~]# su - mongodb
创建目录和文件
[mongodb@node1 ~]$ mkdir /data/{config,shard1,shard2,shard3,mongos,logs,configsvr,keyfile} -pv[mongodb@node1 ~]$ touch /data/keyfile/zxl[mongodb@node1 ~]$ touch /data/logs/shard{1..3}.log[mongodb@node1 ~]$ touch /data/logs/{configsvr,mongos}.log[mongodb@node1 ~]$ touch /data/config/shard{1..3}.conf[mongodb@node1 ~]$ touch /data/config/{configsvr,mongos}.conf
下载mongodb
[mongodb@node1 ~]$ wget https://fastdl.mongodb.org/linux/mongodb-linux-x86_64-rhel62-3.2.3.tgz[mongodb@node3 ~]$ tar fxz mongodb-linux-x86_64-rhel62-3.2.3.tgz -C /data[mongodb@node3 ~]$ ln -s /data/mongodb-linux-x86_64-rhel62-3.2.3 /data/mongodb
配置mongodb环境变量
[mongodb@node1 ~]$ echo "export PATH=$PATH:/data/mongodb/bin" >> ~/.bash_profile[mongodb@node1 data]$ source ~/.bash_profile
shard1.conf配置文件内容如下:
[mongodb@node1 ~]$ cat /data/config/shard1.conf
systemLog: destination: file path: /data/logs/shard1.log logAppend: trueprocessManagement: fork: true pidFilePath: "/data/shard1/shard1.pid"net: port: 10001storage: dbPath: "/data/shard1" engine: wiredTiger journal: enabled: true directoryPerDB: trueoperationProfiling: slowOpThresholdMs: 10 mode: "slowOp"
#security:# keyFile: "/data/keyfile/zxl"# clusterAuthMode: "keyFile"replication: oplogSizeMB: 50 replSetName: "shard1_zxl" secondaryIndexPrefetch: "all"
shard2.conf配置文件内容如下:
[mongodb@node1 ~]$ cat /data/config/shard2.conf
systemLog: destination: file path: /data/logs/shard2.log logAppend: trueprocessManagement: fork: true pidFilePath: "/data/shard2/shard2.pid"net: port: 10002storage: dbPath: "/data/shard2" engine: wiredTiger journal: enabled: true directoryPerDB: trueoperationProfiling: slowOpThresholdMs: 10 mode: "slowOp"
#security:# keyFile: "/data/keyfile/zxl"# clusterAuthMode: "keyFile"replication: oplogSizeMB: 50 replSetName: "shard2_zxl" secondaryIndexPrefetch: "all"
shard3.conf配置文件内容如下:
[mongodb@node1 ~]$ cat /data/config/shard3.conf
systemLog: destination: file path: /data/logs/shard3.log logAppend: trueprocessManagement: fork: true pidFilePath: "/data/shard3/shard3.pid"net: port: 10003storage: dbPath: "/data/shard3" engine: wiredTiger journal: enabled: true directoryPerDB: trueoperationProfiling: slowOpThresholdMs: 10 mode: "slowOp"#security:# keyFile: "/data/keyfile/zxl"# clusterAuthMode: "keyFile"replication: oplogSizeMB: 50 replSetName: "shard3_zxl" secondaryIndexPrefetch: "all"
configsvr.conf配置文件内容如下:
[mongodb@node1 ~]$ cat /data/config/configsvr.conf
systemLog: destination: file path: /data/logs/configsvr.log logAppend: trueprocessManagement: fork: true pidFilePath: "/data/configsvr/configsvr.pid"net: port: 10004storage: dbPath: "/data/configsvr" engine: wiredTiger journal: enabled: true#security:# keyFile: "/data/keyfile/zxl"# clusterAuthMode: "keyFile"sharding: clusterRole: configsvr
mongos.conf配置文件内容如下:
[mongodb@node3 ~]$ cat /data/config/mongos.conf
systemLog: destination: file path: /data/logs/mongos.log logAppend: trueprocessManagement: fork: true pidFilePath: /data/mongos/mongos.pidnet: port: 10005sharding: configDB: 192.168.42.41:10004,192.168.42.42:10004,192.168.42.43:10004#security:# keyFile: "/data/keyfile/zxl"# clusterAuthMode: "keyFile"
注:以上操作只是在node1机器上操作,请把上面这些操作步骤在另外2台机器操作一下,包括创建用户创建目录文件以及安装mongodb等,以及文件拷贝到node2、node3对应的目录下,拷贝之后查看一下文件的属主属组是否为mongodb。
启动各个机器节点的mongod,shard1、shard2、shard3
[mongodb@node1 ~]$ mongod -f /data/config/shard1.confmongod: /usr/lib64/libcrypto.so.10: no version information available (required by mmongod: /usr/lib64/libcrypto.so.10: no version information available (required by mmongod: /usr/lib64/libssl.so.10: no version information available (required by mongmongod: relocation error: mongod: symbol TLSv1_1_client_method, version libssl.so.1n file libssl.so.10 with link time reference
注:无法启动,看到相应的提示后
解决:安装openssl即可,三台机器均安装openssl-devel,redhat6.6基本不存在这种情况,centos可能会遇到。
[mongodb@node1 ~]$ su - rootPassword: [root@node1 ~]# yum install openssl-devel -y
再次切换mongodb用户启动三台机器上的mongod,shard1、shard2、shard3
[mongodb@node1 ~]$ mongod -f /data/config/shard1.confabout to fork child process, waiting until server is ready for connections.forked process: 1737child process started successfully, parent exiting[mongodb@node1 ~]$ mongod -f /data/config/shard2.confabout to fork child process, waiting until server is ready for connections.forked process: 1760child process started successfully, parent exiting[mongodb@node1 ~]$ mongod -f /data/config/shard3.confabout to fork child process, waiting until server is ready for connections.forked process: 1783child process started successfully, parent exiting
注意:如果这一步报错,可以看他的提示信息,或者/data/logs/shard{1..3}.log的日志。
进入node1机器上的mongod:10001登录
[mongodb@node1 ~]$ mongo --port 10001MongoDB shell version: 3.2.3connecting to: 127.0.0.1:10001/testWelcome to the MongoDB shell.For interactive help, type "help".For more comprehensive documentation, seehttp://docs.mongodb.org/Questions? Try the support grouphttp://groups.google.com/group/mongodb-userServer has startup warnings: 2016-03-08T13:28:18.508+0800 I CONTROL [initandlisten] 2016-03-08T13:28:18.508+0800 I CONTROL [initandlisten] ** WARNING: /sys/kernel/mm/epage/enabled is 'always'.2016-03-08T13:28:18.508+0800 I CONTROL [initandlisten] ** We suggest settin2016-03-08T13:28:18.508+0800 I CONTROL [initandlisten] 2016-03-08T13:28:18.508+0800 I CONTROL [initandlisten] ** WARNING: /sys/kernel/mm/epage/defrag is 'always'.2016-03-08T13:28:18.508+0800 I CONTROL [initandlisten] ** We suggest settin2016-03-08T13:28:18.508+0800 I CONTROL [initandlisten]
注:提示warning......
解决:在三台机器上均操作一下内容即可
[mongodb@node2 config]$ su - rootPassword[root@node1 ~]# echo never > /sys/kernel/mm/transparent_hugepage/enabled[root@node1 ~]# echo never > /sys/kernel/mm/transparent_hugepage/defrag
一定要关闭三台机器上的mongod实例,然后再次启动三台机器上mongod实例,否则不生效。
[mongodb@node1 ~]$ netstat -ntpl|grep mongo|awk '{print $NF}'|awk -F'/' '{print $1}'|xargs kill [mongodb@node1 ~]$ mongod -f /data/config/shard1.conf[mongodb@node1 ~]$ mongod -f /data/config/shard2.conf[mongodb@node1 ~]$ mongod -f /data/config/shard3.conf
看端口是否存在:[mongodb@elk_node1 ~]$ ss -atunlp|grep mongtcp LISTEN 0 128 *:10001 *:* users:(("mongod",2630,6))tcp LISTEN 0 128 *:10002 *:* users:(("mongod",2654,6))tcp LISTEN 0 128 *:10003 *:* users:(("mongod",2678,6))
配置复制集
node1机器上操作配置复制集
[mongodb@node1 config]$ mongo --port 10001MongoDB shell version: 3.2.3connecting to: 127.0.0.1:10001/test> use adminswitched to db admin
> config = { _id:"shard1_zxl", members:[... ... {_id:0,host:"192.168.42.41:10001"},... ... {_id:1,host:"192.168.42.42:10001"},... ... {_id:2,host:"192.168.42.43:10001",arbiterOnly:true}... ... ]... ... }
{"_id" : "shard1_zxl","members" : [{"_id" : 0,"host" : "192.168.42.41:10001"},{"_id" : 1,"host" : "192.168.42.42:10001"},{"_id" : 2,"host" : "192.168.42.43:10001","arbiterOnly" : true}]}
> rs.initiate(conconfig connect( connectionURLTheSame( constructor> rs.initiate(config){ "ok" : 1 }
node2机器上操作配置复制集
[mongodb@node2 config]$ mongo --port 10002MongoDB shell version: 3.2.3connecting to: 127.0.0.1:10002/testWelcome to the MongoDB shell.For interactive help, type "help".For more comprehensive documentation, seehttp://docs.mongodb.org/Questions? Try the support grouphttp://groups.google.com/group/mongodb-user> use adminswitched to db admin
> config = { _id:"shard2_zxl", members:[... ... {_id:0,host:"192.168.42.42:10002"},... ... {_id:1,host:"192.168.42.43:10002"},... ... {_id:2,host:"192.168.42.41:10002",arbiterOnly:true}... ... ]... ... }{"_id" : "shard2_zxl","members" : [{"_id" : 0,"host" : "192.168.42.42:10002"},{"_id" : 1,"host" : "192.168.42.43:10002"},{"_id" : 2,"host" : "192.168.42.41:10002","arbiterOnly" : true}]}
> rs.initiate(config){ "ok" : 1 }
node3机器上操作配置复制集
[mongodb@node3 config]$ mongo --port 10003MongoDB shell version: 3.2.3connecting to: 127.0.0.1:10003/testWelcome to the MongoDB shell.For interactive help, type "help".For more comprehensive documentation, seehttp://docs.mongodb.org/Questions? Try the support grouphttp://groups.google.com/group/mongodb-user> use adminswitched to db admin
> config = {_id:"shard3_zxl", members:[... ... {_id:0,host:"192.168.42.43:10003"},... ... {_id:1,host:"192.168.42.41:10003"},... ... {_id:2,host:"192.168.42.42:10003",arbiterOnly:true}... ... ]... ... }{"_id" : "shard3_zxl","members" : [{"_id" : 0,"host" : "192.168.42.43:10003"},{"_id" : 1,"host" : "192.168.42.41:10003"},{"_id" : 2,"host" : "192.168.42.42:10003","arbiterOnly" : true}]}
> rs.initiate(config){ "ok" : 1 }
注:以上是配置rs复制集,相关命令如:rs.status(),查看各个复制集的状况
启动三台机器上的configsvr和mongos节点
[mongodb@node1 logs]$ mongod -f /data/config/configsvr.confabout to fork child process, waiting until server is ready for connections.forked process: 6317child process started successfully, parent exiting
[mongodb@node1 logs]$ mongos -f /data/config/mongos.confabout to fork child process, waiting until server is ready for connections.forked process: 6345child process started successfully, parent exiting
查看端口是否启动:
[mongodb@elk_node1 ~]$ ss -atunlp|grep mongtcp LISTEN 0 128 *:10001 *:* users:(("mongod",2630,6))tcp LISTEN 0 128 *:10002 *:* users:(("mongod",2654,6))tcp LISTEN 0 128 *:10003 *:* users:(("mongod",2678,6))tcp LISTEN 0 128 *:10004 *:* users:(("mongod",3053,6))tcp LISTEN 0 128 *:10005 *:* users:(("mongos",3157,21))
配置shard分片
在node1机器上配置shard分片
[mongodb@node1 config]$ mongo --port 10005MongoDB shell version: 3.2.3connecting to: 127.0.0.1:10005/testmongos> use adminswitched to db admin
mongos> db.runCommand({addshard:"shard1_zxl/192.168.42.41:10001,192.168.42.42:10001,192.168.42.43:10001"});{ "shardAdded" : "shard1_zxl", "ok" : 1 }
mongos> db.runCommand({addshard:"shard2_zxl/192.168.42.41:10002,192.168.42.42:10002,192.168.42.43:10002"});{ "shardAdded" : "shard2_zxl", "ok" : 1 }
mongos> db.runCommand({addshard:"shard3_zxl/192.168.42.41:10003,192.168.42.42:10003,192.168.42.43:10003"});{ "shardAdded" : "shard3_zxl", "ok" : 1 }
查看shard信息
mongos> sh.status()--- Sharding Status --- sharding version: {"_id" : 1,"minCompatibleVersion" : 5,"currentVersion" : 6,"clusterId" : ObjectId("56de6f4176b47beaa9c75e9d")} shards:{ "_id" : "shard1_zxl", "host" : "shard1_zxl/192.168.42.41:10001,192.168.42.42:10001" }{ "_id" : "shard2_zxl", "host" : "shard2_zxl/192.168.42.42:10002,192.168.42.43:10002" }{ "_id" : "shard3_zxl", "host" : "shard3_zxl/192.168.42.41:10003,192.168.42.43:10003" } active mongoses:"3.2.3" : 3 balancer:Currently enabled: yesCurrently running: noFailed balancer rounds in last 5 attempts: 0Migration Results for the last 24 hours: No recent migrations databases:
查看分片状态
mongos> db.runCommand( {listshards : 1 } ){"shards" : [{"_id" : "shard1_zxl","host" : "shard1_zxl/192.168.42.41:10001,192.168.42.42:10001"},{"_id" : "shard2_zxl","host" : "shard2_zxl/192.168.42.42:10002,192.168.42.43:10002"},{"_id" : "shard3_zxl","host" : "shard3_zxl/192.168.42.41:10003,192.168.42.43:10003"}],"ok" : 1}
启用shard分片的库名字为'zxl',即为库
mongos> sh.enableSharding("zxl"){ "ok" : 1 }
设置集合的名字以及字段,默认自动建立索引,zxl库,haha集合
mongos> sh.shardCollection("zxl.haha",{age: 1, name: 1}){ "collectionsharded" : "zxl.haha", "ok" : 1 }
模拟在haha集合中插入10000数据
mongos> for (i=1;i<=10000;i++) db.haha.insert({name: "user"+i, age: (i0)})WriteResult({ "nInserted" : 1 })
可以使用上面mongos> sh.status()命令查看各个shard分片情况,以上就是复制集和shard分片搭建完成。
结果如下:
mongos> sh.status()--- Sharding Status --- sharding version: {"_id" : 1,"minCompatibleVersion" : 5,"currentVersion" : 6,"clusterId" : ObjectId("56de6f4176b47beaa9c75e9d")} shards:{ "_id" : "shard1_zxl", "host" : "shard1_zxl/192.168.42.41:10001,192.168.42.42:10001" }{ "_id" : "shard2_zxl", "host" : "shard2_zxl/192.168.42.42:10002,192.168.42.43:10002" }{ "_id" : "shard3_zxl", "host" : "shard3_zxl/192.168.42.41:10003,192.168.42.43:10003" } active mongoses:"3.2.3" : 3 balancer:Currently enabled: yesCurrently running: noFailed balancer rounds in last 5 attempts: 0Migration Results for the last 24 hours: 2 : Success databases:{ "_id" : "zxl", "primary" : "shard3_zxl", "partitioned" : true }zxl.haha
shard key: { "age" : 1, "name" : 1 }unique: falsebalancing: truechunks:shard1_zxl1shard2_zxl1shard3_zxl1{ "age" : { "$minKey" : 1 }, "name" : { "$minKey" : 1 } } -->> { "age" : 2, "name" : "user2" } on : shard1_zxl Timestamp(2, 0) { "age" : 2, "name" : "user2" } -->> { "age" : 22, "name" : "user22" } on : shard2_zxl Timestamp(3, 0) { "age" : 22, "name" : "user22" } -->> { "age" : { "$maxKey" : 1 }, "name" : { "$maxKey" : 1 } } on : shard3_zxl Timestamp(3, 1)
以上就是mongodb3.2复制集和shard分片搭建就此完成.有什么疑问可以给我留言,看到第一时间解决。