The latest 4.2.7 version of MongoDB three-slice cluster modification IP practice exercise

Background
Re-networking requires changes to the IP of the existing MongoDB sharded cluster server, so the IP of the MongoDB sharded cluster also needs to be updated accordingly, and the IP modification of the MongoDB sharded cluster cannot be done simply by configuration , It takes some toss to update normally, here is a record of the IP modification process of the entire MongoDB cluster, I hope it will be helpful to those with the same needs.

The original cluster configuration is as follows

. The port allocation here is as follows: The

corresponding relationship between IP modification is as follows:

Change IP steps

The upgrade process is to upgrade the Config service first , then upgrade the three Shard configurations, and finally modify mongos the route of one. Of course, you need to make sure that the MongDB sharded cluster is stopped before starting.
The steps to stop are also very simple. ps -ef | grep mongo Find the corresponding thread id and kill drop it.

Modify the IP of the config node

Use the following command to standlone start a Config service in mode

/usr/mongod/bin/mongod --port 21000 --dbpath=/usr/mongod/data

Open a new shell and connect to the config service

/usr/mongod/bin/mongo --port 21000

Go to the local database to view the replset collection, and modify the correspondinghost

use local
db.system.replset.find()
cfg = db.system.replset.findOne({_id: 'configs'})
cfg.members[0].host = "172.168.7.11:21000"
cfg.members[1].host = "172.168.7.16:21000"
cfg.members[2].host = "172.168.7.21:21000"
db.system.replset.update({_id: 'configs'}, cfg)
db.system.replset.find()

The last line above is to find out the modified content in order to check whether the modification is successful.
Follow the same steps to standlone start the other two config nodes in mode and modify the information of the other two config service nodes.
Stop these three standlone config service nodes started in mode
Start the three config nodes according to the normal cluster mode respectively

/usr/mongod/bin/mongod -f /usr/mongod/conf/config.conf

Seeing the following shows that the election was successful.
Log in to the primary node and modify the shard information

use config

cfg=db.shards.findOne({_id:'shard1'})
cfg.host="shard1/172.168.7.11:27001,172.168.7.16:27001"
db.shards.update({_id:'shard1'},cfg)
cfg = db.shards.findOne({_id: 'shard2'})

cfg.host = "shard2/172.168.7.16:27002,172.168.7.21:27002"
db.shards.update({_id: 'shard2'}, cfg)

cfg = db.shards.findOne({_id: 'shard3'})
cfg.host = "shard3/172.168.7.11:27003,172.168.7.21:27003"
db.shards.update({_id: 'shard3'}, cfg)

So far, the modification of the config service node is completed, and then the modification of the shard node

Modify the IP information of the shard fragment

standlone Start shard 1 in mode, then log in

/usr/mongod/bin/mongod --port 27001 --dbpath=/usr/mongod/shard1/
/usr/mongod/bin/mongo --port 27001

Switch to admin the database version collection of view the config configuration information, if _id:shardIdentity a data update is performed next, of course, if there _id:minOpTimeRecovery is data need to be updated

use admin
db.system.version.find()
db.system.version.update({"_id" : "shardIdentity"},{"$set":{"configsvrConnectionString" : "configs/172.168.7.11:21000,172.168.7.16:21000,172.168.7.21:21000"}})
db.system.version.update({"_id" : "minOpTimeRecovery"},{"$set":{"configsvrConnectionString" : "configs/172.168.7.11:21000,172.168.7.16:21000,172.168.7.21:21000"}})

Modify the configuration information of the replication set, the configuration information of the replication set is saved local in the system.replset collection of the library

use local
cfg = db.system.replset.findOne({_id: 'shard1'})
cfg.members[0].host="172.168.7.11:27003"
cfg.members[1].host="172.168.7.16:27003"
cfg.members[2].host="172.168.7.21:27003"
db.system.replset.update({_id:'shard3'},cfg)
db.system.replset.findOne({_id: 'shard3'})

Repeat the above steps to change the ip of shaed2 and shard3 respectively
Finally, start your three shards according to the cluster mode.

/usr/mongod/bin/mongod -f /usr/mongod/conf/shard1.conf

Modify the routing information of the mongos service

Open the configuration file of mongos with vim and modify the corresponding configserver configuration item to the latest configderver address

vim /usr/mongod/conf/mongos.conf

Start the mongos service in cluster mode

/usr/mongod/bin/mongos -f /usr/mongod/conf/mongos.conf

Just login and verify

/usr/mongod/bin/mongo --port 20000

At this point, the IP modification of the MongoDB sharded cluster is over.