mongo 对删除节点remove OTHER

素颜马尾好姑娘i 2022-02-20 09:54 466阅读 0赞

在复制集之中，我们通常会有删除复制集节点的操作

其中一个很典型的场景就是，海量数据的跨机房无缝迁移。例如，A机房  有一主一从两节点复制集，现在要把数据迁移到  B机房，并做到过程耗时很短，同时迁移失败及时回滚还原直接的连接配置。

一般会将B机房一个节点作为A机房复制集的从节点，直到同步趋于一致之后，再将连接改成B机房的，但是此过程会有如下问题：

stage001:PRIMARY> rs.remove("10.18.19.207:27019")

\{ "ok" : 1 \}

stage001:OTHER> rs.slaveOk();

stage001:OTHER> show dbs

admin 0.000GB

braavos 2.385GB

cms 3.997GB

loanapp 0.000GB

local 4.178GB

morpheus 0.106GB

scportal 0.000GB

sns 0.001GB

stage001:OTHER> rs.status()

"state" : 10,

"stateStr" : "REMOVED",

"uptime" : 10444617,

"optime" : \{

"ts" : Timestamp(1539832013, 1),

"t" : NumberLong(27)

\},

"optimeDate" : ISODate("2018-10-18T03:06:53Z"),

"ok" : 0,

"errmsg" : "Our replica set config is invalid or we are not a member of it",

"code" : 93,

"codeName" : "InvalidReplicaSetConfig"

以上是已删除节点的状态，此状态是无法读写，做复制集的。

查看本机配置信息

stage001:OTHER> use local

switched to db local

stage001:OTHER> show tables;

clustermanager

oplog.rs

replset.election

replset.minvalid

startup\_log

system.replset

stage001:OTHER> db.system.replset.find()

\{ "\_id" : "stage001", "version" : 2, "protocolVersion" : NumberLong(1), "members" : \[ \{ "\_id" : 1, "host" : "10.18.19.207:27019", "arbiterOnly" : false, "buildIndexes" : true, "hidden" : false, "priority" : 1, "tags" : \{ \}, "slaveDelay" : NumberLong(0), "votes" : 1 \}, \{ "\_id" : 2, "host" : "10.18.19.208:27019", "arbiterOnly" : false, "buildIndexes" : true, "hidden" : false, "priority" : 1, "tags" : \{ \}, "slaveDelay" : NumberLong(0), "votes" : 1 \} \], "settings" : \{ "chainingAllowed" : true, "heartbeatIntervalMillis" : 2000, "heartbeatTimeoutSecs" : 10, "electionTimeoutMillis" : 10000, "catchUpTimeoutMillis" : 60000, "getLastErrorModes" : \{ \}, "getLastErrorDefaults" : \{ "w" : 1, "wtimeout" : 0 \}, "replicaSetId" : ObjectId("5ab8ca2134014d920ae64248") \} \}

究其原因，是因为此时人保有之前的复制集配置信息，此时需要删除此信息

但是在local系统库下面的  system.replset  就连root权限的用户都无法删除，此时我们需要一个特殊权限的用户，即如下

主库创建内部权限角色用户

stage001:PRIMARY> db.createUser(\{user: "fafa",pwd: "fafa123",roles: \[ \{ role: "\_\_system", db: "admin" \} \]\})

Successfully added user: \{

"user" : "fafa",

"roles" : \[

"role" : "\_\_system",

"db" : "admin"

stage001:OTHER> use admin

switched to db admin

stage001:OTHER> db.auth('fafa','fafa123')

stage001:OTHER> use local

switched to db local

stage001:OTHER> rs.slaveOk();

stage001:OTHER> show tables;

clustermanager

oplog.rs

replset.election

replset.minvalid

startup\_log

system.replset

stage001:OTHER> db.system.replset.find()

\{ "\_id" : "stage001", "version" : 3, "protocolVersion" : NumberLong(1), "members" : \[ \{ "\_id" : 2, "host" : "10.18.19.208:27019", "arbiterOnly" : false, "buildIndexes" : true, "hidden" : false, "priority" : 1, "tags" : \{ \}, "slaveDelay" : NumberLong(0), "votes" : 1 \} \], "settings" : \{ "chainingAllowed" : true, "heartbeatIntervalMillis" : 2000, "heartbeatTimeoutSecs" : 10, "electionTimeoutMillis" : 10000, "catchUpTimeoutMillis" : 60000, "getLastErrorModes" : \{ \}, "getLastErrorDefaults" : \{ "w" : 1, "wtimeout" : 0 \}, "replicaSetId" : ObjectId("5ab8ca2134014d920ae64248") \} \}

stage001:OTHER> db.system.replset.remove(\{ "\_id" : "stage001", "version" : 3, "protocolVersion" : NumberLong(1), "members" : \[ \{ "\_id" : 2, "host" : "10.18.19.208:27019", "arbiterOnly" : false, "buildIndexes" : true, "hidden" : false, "priority" : 1, "tags" : \{ \}, "slaveDelay" : NumberLong(0), "votes" : 1 \} \], "settings" : \{ "chainingAllowed" : true, "heartbeatIntervalMillis" : 2000, "heartbeatTimeoutSecs" : 10, "electionTimeoutMillis" : 10000, "catchUpTimeoutMillis" : 60000, "getLastErrorModes" : \{ \}, "getLastErrorDefaults" : \{ "w" : 1, "wtimeout" : 0 \}, "replicaSetId" : ObjectId("5ab8ca2134014d920ae64248") \} \})

WriteResult(\{ "nRemoved" : 1 \})

此时，B节点已变成一个单节点的状态，可以变成任何的复制集或者集群架构了