MySQL DBA
分类： NOSQL
2013-03-07 14:03:43
mongodb replica set config
参考：
1. rs.conf() replica set 相关的配置类，具有的属性：
a) _id, replica set 名称。
b) members。各节点相关属性（id,host,priority,tag,slaveDelay…）。
c) settings。配置信息。
2. rs 操作类：
rs.status() 返回当前replica set 的状态信息。每个member的host，health，role，optime等信息
rs.initate() 初始化配置一个复制集，加入参数则按照参数设定配置。
rs.conf() 返回当前复制集的配置信息。
rs.reconfig() 在线更新复制集配置信息，如对member 的host等属性的更改。
rs.add() 添加节点。参数可以是host：port，也可是完整的结构体。
rs.stepDown() primary 主动降级为slave。
rs.freeze() 设置节点一定时间内不能被选为primary，在切换的时候人工干预。
rs.remove() 移除节点。
db.isMaster() 返回整个复制集hosts，以及角色的分配情况。在rs.status()也能看到。
搭建步骤：
1. 启动实例
mongod --port 27018 --dbpath /data/mongodb/rs-1 --replSet rs0 &
mongod --port 27019 --dbpath /data/mongodb/rs-2 --replSet rs0 &
mongod --port 27020 --dbpath /data/mongodb/rs-3 --replSet rs0 &
2. 配置节点
ubuntu@os-cloud-12:~$ mongo localhost:27018
MongoDB shell version: 2.0.4
connecting to: localhost:27018/test
> rs.status()
{
"startupStatus" : 3,
"info" : "run rs.initiate(...) if not yet done for the set",
"errmsg" : "can't get local.system.replset config from self or any seed (EMPTYCONFIG)",
"ok" : 0
}
#使用默认配置初始化。
> rs.initiate()
{
"info2" : "no configuration explicitly specified -- making one",
"me" : "os-cloud-12:27018",
"info" : "Config now saved locally. Should come online in about a minute.",
"ok" : 1
}
> rs.conf()
{
"_id" : "rs0",
"version" : 1,
"members" : [
{
"_id" : 0,
"host" : "os-cloud-12:27018"
}
]
}
#初始化后，默认当前host为primary，继续添加节点
PRIMARY> rs.add("os-cloud-12:27019")
{ "ok" : 1 }
PRIMARY> rs.add("os-cloud-12:27020")
{ "ok" : 1 }
#查看配置情况和当前状态
PRIMARY> rs.conf()
{
"_id" : "rs0",
"version" : 3,
"members" : [
{
"_id" : 0,
"host" : "os-cloud-12:27018"
},
{
"_id" : 1,
"host" : "os-cloud-12:27019"
},
{
"_id" : 2,
"host" : "os-cloud-12:27020"
}
]
}
PRIMARY> rs.status()
{
"set" : "rs0",
"date" : ISODate("2012-11-24T03:01:54Z"),
"myState" : 1,
"syncingTo" : "os-cloud-12:27018",
"members" : [
{
"_id" : 0,
"name" : "os-cloud-12:27018",
"health" : 1,
"state" : 2,
"stateStr" : "SECONDARY",
"uptime" : 60854,
"optime" : {
"t" : 1353665255000,
"i" : 1
},
"optimeDate" : ISODate("2012-11-23T10:07:35Z"),
"lastHeartbeat" : ISODate("2012-11-24T03:01:53Z"),
"pingMs" : 0
},
{
"_id" : 1,
"name" : "os-cloud-12:27019",
"health" : 1,
"state" : 2,
"stateStr" : "SECONDARY",
"uptime" : 60854,
"optime" : {
"t" : 1353665255000,
"i" : 1
},
"optimeDate" : ISODate("2012-11-23T10:07:35Z"),
"lastHeartbeat" : ISODate("2012-11-24T03:01:53Z"),
"pingMs" : 0
},
{
"_id" : 2,
"name" : "os-cloud-12:27020",
"health" : 1,
"state" : 1,
"stateStr" : "PRIMARY",
"optime" : {
"t" : 1353665255000,
"i" : 1
},
"optimeDate" : ISODate("2012-11-23T10:07:35Z"),
"self" : true
}
],
"ok" : 1
}
3. primary 切换。
PRIMARY> rs.stepDown()
Fri Nov 23 18:12:31 DBClientCursor::init call() failed
Fri Nov 23 18:12:31 query failed : admin.$cmd { replSetStepDown: 60.0 } to: localhost:27018
Fri Nov 23 18:12:31 Error: error doing query: failed shell/collection.js:151
Fri Nov 23 18:12:31 trying reconnect to localhost:27018
Fri Nov 23 18:12:31 reconnect localhost:27018 ok
SECONDARY> rs.status()
4. member 属性的修改。
a) 重启需要修改的mongodb 节点，以新的配置加入复制集
b) 在primary中修改其属性值（rs.reconfig()）
选举策略：
1 get maxLocalOpOrdinal from each server.
2 if a majority of servers are not up (from this server's POV), remain in Secondary mode and stop.
3 if the last op time seems very old, stop and await human intervention.
4 else, using a consensus protocol, pick the server with the highest maxLocalOpOrdinal as the Primary.
http://blog.nosqlfan.com/html/2523.html
