Shulou(Shulou.com)06/01 Report--

CentOS6 platform installation MongoDB3.2 copy set

One, three machines are all installed and deployed mongod

1. Download the installation package and extract the tgz

Curl-O https://fastdl.mongodb.org/linux/mongodb-linux-x86_64-3.2.10.tgz

Tar xf mongodb-linux-x86_64-3.2.10.tgz

Mv mongodb-linux-x86_64-3.2.10 / usr/local/mongodb

2. Add the executable file to the PATH path:

Export PATH=/usr/local/mongodb/bin:$PATH

Echo 'export PATH=/usr/local/mongodb/bin:$PATH' > / etc/profile.d/mongodb.sh

3. Close large page memory

Echo never > / sys/kernel/mm/transparent_hugepage/enabled

Echo never > / sys/kernel/mm/transparent_hugepage/defrag

Echo "if test-f / sys/kernel/mm/transparent_hugepage/enabled;then

Echo never > / sys/kernel/mm/transparent_hugepage/enabled

Fi

If test-f / sys/kernel/mm/transparent_hugepage/defrag;then

Echo never > / sys/kernel/mm/transparent_hugepage/defrag

Fi "> > / etc/rc.d/rc.local

4, limit

Ulimit-f unlimited

Ulimit-t unlimited

Ulimit-v unlimited

Ulimit-n 64000

Ulimit-m unlimited

Ulimit-u 64000

5. Create a database directory

Mkdir-p / data0/mongodb/shardA/data

Mkdir-p / data0/logs/mongodb/

6, run the MongoDB service

(1) write configuration files

Cat / data0/mongodb/shardA/mongo27017.conf

# mongodb shardA 27017

ReplSet=shardA

Dbpath=/data0/mongodb/shardA/data

Logpath=/data0/logs/mongodb/27017.log

# bind=192.168.9.168127.0.0.1

Port=27017

Fork=true

MaxConns=20000

Logappend=true

Smallfiles=true

(2) start

/ usr/local/mongodb/bin/mongod-f / data0/mongodb/shardA/mongo27017.conf

# start it in other ways

The command line specifies dbpath and port

/ usr/local/mongodb/bin/mongod-- dbpath=/data0/mongodb/shardA-- port 27017 & # default port is 27017

The command line specifies dbpath and port and launches the web interface

/ usr/local/mongodb/bin/mongod-- dbpath=/data0/mongodb/shardA-- port 27017-- rest & # web default port is 28017

7, check to see if the startup is normal

Netstat-lnptu | grep 27017

Ps auxf | grep mongod | grep-v grep

MongoDB backend manages shell

/ usr/local/mongodb/bin/mongo 127.0.0.1:27017/test

/ usr/local/mongodb/bin/mongo-- port 27017

Second, configure the replica set

1Bing # connects to a mongo

# / usr/local/mongodb/bin/mongo 192.168.9.168:27017

MongoDB shell version: 3.2.10

2 priority # configuration 168master, 176slave, 225arbitration node, priority maximum master node, arbiterOnly only as arbitration

> cfg= {_ id: "shardA", members: [{_ id:0,host:'192.168.9.168:27017',priority:2}, {_ id:1,host:'192.168.9.176:27017',priority:1}, {_ id:2,host:'192.168.9.225:27017',arbiterOnly:true}]}

{

"_ id": "shardA"

"members": [

{

"_ id": 0

"host": "192.168.9.168 purl 27017"

"priority": 2

}

{

"_ id": 1

"host": "192.168.9.176VR 27017"

"priority": 1

}

{

"_ id": 2

"host": "192.168.9.225 purl 27017"

"arbiterOnly": true

}

]

}

# initialization

> rs.initiate (cfg)

{"ok": 1}

# you can also add it manually

> rs.initiate ()

> rs.add ("192.168.9.176 purl 27017")

> rs.add ("192.168.9.225 virtual 27017", {arbiterOnly:true})

{"ok": 1}

# deleting a node

> rs.remove ()

3, check the status

# View configuration content

> rs.conf ()

# View status information

> rs.status ()

# check whether it is the primary node

> db.isMaster ()

4, the master node inserts data

Mongo 192.168.9.176:27017

> show dbs

> use Atest

> for (iSuppli show tables)

> db.coltest.count ()

> db.coltest.find ()

Mongo 192.168.9.168:27017

> rs.slaveOk ()

> show dbs

> show tables

> db.coltest.count ()

> db.coltest.find ()

ShardA:SECONDARY > db.coltest.insert ({"count": "10003"})

WriteResult ({"writeError": {"code": 10107, "errmsg": "not master"}})

Third, fail-over testing

1, close the primary node

Mongo 192.168.9.176:27017

ShardA:PRIMARY > use admin

Switched to db admin

ShardA:PRIMARY > db.shutdownServer ()

Server should be down...

2017-11-30T14:13:23.223+0800 I NETWORK [thread1] trying reconnect to 127.0.0.1 thread1 27017 (127.0.0.1) failed

2017-11-30T14:13:24.116+0800 I NETWORK [thread1] Socket recv () errno:104 Connection reset by peer 127.0.0.1 purl 27017

2017-11-30T14:13:24.116+0800 I NETWORK [thread1] SocketException: remote: (NONE): 0 error: 9001 socket exception [RECV_ERROR] server [127.0.0.1 virtual 27017]

2017-11-30T14:13:24.116+0800 I NETWORK [thread1] reconnect 127.0.0.1 thread1 27017 (127.0.0.1) failed failed

2017-11-30T14:13:24.118+0800 I NETWORK [thread1] trying reconnect to 127.0.0.1 thread1 27017 (127.0.0.1) failed

2017-11-30T14:13:24.118+0800 W NETWORK [thread1] Failed to connect to 127.0.0.1 thread1 27017, reason: errno:111 Connection refused

2017-11-30T14:13:24.118+0800 I NETWORK [thread1] reconnect 127.0.0.1 thread1 27017 (127.0.0.1) failed failed

>

2. Check that the standby node will become the primary node.

ShardA:SECONDARY >

ShardA:PRIMARY > rs.status ()

{

"set": "shardA"

"date": ISODate ("2017-11-30T06:13:57.404Z")

"myState": 1

"term": NumberLong (3)

"heartbeatIntervalMillis": NumberLong (2000)

"members": [

{

"_ id": 0

"name": "192.168.9.168 purl 27017"

"health": 1

"state": 1

"stateStr": "PRIMARY"

"uptime": 74322

"optime": {

Ts: Timestamp (1512022413, 1)

"t": NumberLong (3)

}

"optimeDate": ISODate ("2017-11-30T06:13:33Z")

"infoMessage": "could not find member to sync from"

ElectionTime: Timestamp (1512022412, 1)

"electionDate": ISODate ("2017-11-30T06:13:32Z")

"configVersion": 1

"self": true

}

{

"_ id": 1

"name": "192.168.9.176 purl 27017"

"health": 0

"state": 8

"stateStr": "(not reachable/healthy)

"uptime": 0

"optime": {

"ts": Timestamp (0,0)

"t": NumberLong (- 1)

}

OptimeDate: ISODate ("1970-01-01T00:00:00Z")

"lastHeartbeat": ISODate ("2017-11-30T06:13:56.562Z")

"lastHeartbeatRecv": ISODate ("2017-11-30T06:13:21.617Z")

"pingMs": NumberLong (0)

"lastHeartbeatMessage": "Connection refused"

"configVersion":-1

}

{

"_ id": 2

"name": "192.168.9.225 purl 27017"

"health": 1

"state": 7

"stateStr": "ARBITER"

"uptime": 70496

"lastHeartbeat": ISODate ("2017-11-30T06:13:56.543Z")

"lastHeartbeatRecv": ISODate ("2017-11-30T06:13:52.776Z")

"pingMs": NumberLong (0)

"configVersion": 1

}

]

"ok": 1

}

ShardA:PRIMARY >

3. Restart the original master node, and now the master node will cut through it.

ShardA:SECONDARY > db.isMaster ()

{

"hosts": [

"192.168.9.168purl 27017"

"192.168.9.176purl 27017"

]

"arbiters": [

"192.168.9.225purl 27017"

]

"setName": "shardA"

"setVersion": 1

"ismaster": false

"secondary": true

"primary": "192.168.9.176VR 27017"

"me": "192.168.9.168 purl 27017"

"maxBsonObjectSize": 16777216

"maxMessageSizeBytes": 48000000

"maxWriteBatchSize": 1000

"localTime": ISODate ("2017-11-30T06:20:47.355Z")

"maxWireVersion": 4

"minWireVersion": 0

"ok": 1

}

4, close the replica set

Fifth, maintenance related

1. Close the slave node and arbitration node first. Or close the quorum node first and then the slave node. Then: the master node automatically becomes the slave node

2. If the master node is closed first, the slave node automatically becomes the master node, and the new master node cannot be closed until the arbitration node is closed.

During initialization, the active and standby nodes are affected by priority.

Priority: is the priority, default is 1, priority 0 is the passive node, cannot become the active node. If the priority is not 0, the active nodes are selected from the largest to the smallest.

Rs auxiliary function

> rs.help ()

Rs.status () {replSetGetStatus: 1} checks repl set status

Rs.initiate () {replSetInitiate: null} initiates set with default settings

Rs.initiate (cfg) {replSetInitiate: cfg} initiates set with configuration cfg

Rs.conf () get the current configuration object from local.system.replset

Rs.reconfig (cfg) updates the configuration of a running replica set with cfg (disconnects)

Rs.add (hostportstr) add a new member to the set with default attributes (disconnects)

Rs.add (membercfgobj) add a new member to the set with extra attributes (disconnects)

Rs.addArb (hostportstr) add a new member which is arbiterOnly:true (disconnects)

Rs.stepDown ([stepdownSecs, catchUpSecs]) stepdown as primary (disconnects)

Rs.syncFrom (hostportstr) make a secondary sync from the given member

Rs.freeze (secs) make a node ineligible to become primary for the time specified

Rs.remove (hostportstr) remove a host from the replica set (disconnects)

Rs.slaveOk () allow queries on secondary nodes

Rs.printReplicationInfo () check oplog size and time range

Rs.printSlaveReplicationInfo () check replica set members and replication lag

Db.isMaster () check who is primary

Reconfiguration helpers disconnect from the database so the shell will display

An error, even if the command succeeds.

Sixth, design copy set

1. An important concept of replica set is "most", most of which are more than half of the members.

For example, of the replica set of five members, three are unreachable, and two can still work normally, but can not meet most of the requirements, so the primary node cannot be selected. If one is the primary node, it will be switched to the standby node.

The main prevention, 3 networks are unreachable, 3 are communicable, and 2 primary nodes are prevented.

2. The recommended models are:

An odd member of the data center

Two odd members of the data center, putting most of them in one data center

Three data centers, two data centers with an equal number of members, and one data center that determines the primary node

3. Election mechanism

When a backup node is unable to connect with the primary node, it contacts and requests other replica set members to elect themselves.

Other members will do several rational tests:

(1) whether it can communicate with the master node

(2) whether the data of the backup node that wants to be elected as the primary node is up-to-date

(3) are there any other higher priority members who can be elected as the primary node?

Seventh, member configuration item

1. Elect arbitrators

Only participate in the election and do not save the data

(1) there can be at most one arbitrator

(2) if the number of nodes is odd, no arbitrator is needed.

# add an arbitration member

Rs.addArb ("192.168.9.168 purl 27017")

Rs.add ({_ id:2,host:'192.168.9.225:27017',arbiterOnly:true})

2, priority

Priority values range from 0 to 100. The default value is 1. 0 indicates that you can never become a primary node, which is called a passive member.

The higher the priority and the most up-to-date data, it will be elected as the primary node.

3, hide members

The client does not send requests to hidden members, and hidden members do not serve as replication sources

Only members of hidden:true with priority 0 can be hidden

4, delayed backup node

Eight, synchronization

Replication is used to backup data between multiple servers. Using the operation log oplog.

Oplog is a fixed collection in the primary node local database. By querying this collection, the backup node can know the operations that need to be replicated.

1. Initialize synchronization

(1) members will do some preparatory work before recording: select a member as the synchronization source, create an identifier for themselves in local.me, delete all existing databases, and resynchronize

(2) Clone to copy all the records of the synchronous source to the local

(3) enter the first step of oplog synchronization, and all operations in the clone will be recorded in oplog.

(4) the second step of oplog synchronization, which is used to record the operations in the first oplog synchronization

(5) the local data should be exactly the same as the dataset of the master node at a certain point in time, and the index can be created.

(6) if the data of the current node still lags behind the synchronization source, the last step of the oplog synchronization process is to synchronize all operations during the creation of the index to prevent the member from becoming a backup node.

(7) the current member completes the initialization synchronization and switches to the normal synchronization state

Nine, heartbeat.

Members send a heartbeat request to other members every 2s

Member status:

1,STARTUP

The member was in this state when it was first started. Mongodb attempts to load the copy set configuration of the member. After loading successfully, you will enter STARTUP2

2,STARTUP2

The whole initialization synchronization process is in this state.

3,RECOVERING

4,ARBITER

5,DOWN

6,UNKNOWN

7,REMOVED

8,ROLLBACK

9,FATAL

Ten, management

1, start members in stand-alone mode

(1) View command line parameters

> db.serverCmdLineOpts ()

(1) if you want to maintain this server, you can restart it; the replSet option is not applicable when restarting. This will become a stand-alone mongod that can read and write.

(2) the listening port can be changed, and the dbpath value remains unchanged.

(3) after maintenance, restart with the most original parameters will automatically synchronize with other members of the replica set.

2, replica set configuration

(1) create a replica set

# configuration 168master, 176slave, 225arbitration node, priority maximum master node, arbiterOnly only as arbitration

> cfg= {_ id: "shardA", members: [{_ id:0,host:'192.168.9.168:27017',priority:2}, {_ id:1,host:'192.168.9.176:27017',priority:1}, {_ id:2,host:'192.168.9.225:27017',arbiterOnly:true}]}

# initialization

> rs.initiate (cfg)

(2) modify replica set members

A) increase in membership

Rs.add ("192.168.9.169 purl 27017")

Add a specified complex configuration in document form

Rs.add ({_ id:4,host:'192.168.9.169:27017',priority:1})

B) remove members

Rs.remove ("192.168.9.169 purl 27017")

C) modified by rs.config

> var config = rs.config ()

> config.members [0] .host = "192.168.9.179pur27017"

> rs.reconfig (config)

Note:

The _ id field of a member cannot be modified

The priority of the member receiving the rs.reconfig command cannot be set to 0

Cannot change an arbitration member into a non-arbitration member

Cannot change the member of "buildIndexes": false to "buildIndexes": true

3. Create a larger replica set

A copy set can only have a maximum of 12 members, of which only 7 have the right to vote

If you want to create a copy set of more than seven members, only seven members can have the right to vote, and the number of votes of other members needs to be set to 0

> rs.add ({"_ id": 7, "host": server-7:27017, "votes": 0})

4. Modify member status

# the primary node becomes the standby node

# default is 60s

> rs.stepDown (time)

# prevent the election and always be on the standby node

> rs.freeze (time)

# force to enter maintenance mode

> db.adminCommand ({"replSetMaintenance": true})

# exit and enter maintenance mode

> db.adminCommand ({"replSetMaintenance": false})

5, statu

# View synchronization status

> db.printSlaveReplicationInfo ()

# View synchronization sources

> db.adminCommand ({"replSetGetStatus": 1}) .syncingTo

192.168.9.176:27017

# specify replication source

> secondary.adminCommand ({"replSetSyncFrom": "server0:27017"})

# disable replication chain

# replication delay View

# on the primary node

> db.printReplicationInfo ()

# on the standby node

> db.printSlaveReplicationInfo ()

Error report:

ShardA:SECONDARY > show dbs

2017-11-30T13:57:10.376+0800 E QUERY [thread1] Error: listDatabases failed: {"ok": 0, "errmsg": "not master and slaveOk=false", "code": 13435}:

_ getErrorWithCode@src/mongo/shell/utils.js:25:13

Mongo.prototype.getDBs@src/mongo/shell/mongo.js:62:1

ShellHelper.show@src/mongo/shell/utils.js:761:19

ShellHelper@src/mongo/shell/utils.js:651:15

@ (shellhelp2): 1:1

Reason: to protect the application from accidentally connecting to the standby node and reading expired data. The backup node rejects the request by default.

Solution: there is no problem with setting up to read data from the backup node. SalveOk is set for the connection, not for the database.

> rs.slaveOk ()

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.