Docker builds cluster MongoDB implementation steps

Docker builds cluster MongoDB implementation steps

Preface

Due to the needs of the company's business, we plan to build our own MongoDB service. Since MongoDB's cloud database is very expensive, we use the replica set method to build a cluster with three servers, one primary, one secondary, and one arbitration.

Basic Concepts

Replica Set: A replica set is a cluster of MongoDB instances, consisting of a primary server and multiple secondary servers.

  • Master: The master node receives all write operations. The primary node records all changes made to its dataset to its oplog.
  • Secondary: Copies the oplog of the primary node and applies operations to its data set. If the primary node is unavailable, a qualified secondary node will be elected as the new primary node.
  • Arbiter: Load election. When the primary node is unavailable, it will select one of the secondary nodes as the primary node.

Sharding:

Master-slave

  • When running MongoDB 4.0 or later, the following message appears: [main] Master/slave replication is no longer supported. This means that MongoDB 4.0 and later no longer support master-slave replication.

1. Environmental Preparation

Use CentOS 7.6 64bit system to install Docker, Docker-compose, and Docker-Swarm

2. Generate KeyFile

  • MongoDB uses KeyFile authentication. Each MongoDB instance in the replica set uses the KeyFile content as a shared password to authenticate other members. A MongoDB instance can join a replica set only if it has the correct KeyFile.
  • The content of keyFile must be 6 to 1024 characters in length, and the content of KeyFile must be the same for all members of the replica set.
  • One thing to note is that in UNIX systems, KeyFile must not have group permissions or full permissions (that is, the permissions must be set to X00). On Windows, keyFile permissions are not checked.
  • You can use any method to generate the keyFile. For example, the following operation uses openssl to generate a complex random 1024 character string. Then use chmod to modify the file permissions to provide read permissions only to the file owner.
  • This is the official MongoDB recommended way to generate keyFile:
# 400 permissions are required to ensure security, otherwise mongod will report an error when starting openssl rand -base64 756 > mongodb.key
chmod 400 mongodb.key

2. Create a cross-host network

To build a cluster, we must communicate across hosts. To build an Overlay Network, we need to use the Docker Swarm tool. Docker Swarm is a built-in cluster tool for Docker, which can help us deploy services into a cluster of Docker daemons more easily.

Since we want to add Docker to the cluster, we must first have a cluster. We can initialize the cluster through docker swarm init on any Docker instance.

$ sudo docker swarm init

Swarm initialized: current node (t4ydh2o5mwp5io2netepcauyl) is now a manager.

To add a worker to this swarm, run the following command:

  docker swarm join --token SWMTKN-1-4dvxvx4n7magy5zh0g0de0xoues9azekw308jlv6hlvqwpriwy-cb43z26n5jbadk024tx0cqz5r 192.168.1.5:2377

After the cluster is initialized, this Docker instance automatically becomes the cluster management node, and other Docker instances can join the cluster by running the docker swarm join command printed here.

The nodes added to the cluster are ordinary nodes by default. If you want to join the cluster as a management node, you can use the docker swarm join-token command to obtain the management node's join command.

$ sudo docker swarm join-token manager
To add a manager to this swarm, run the following command:

  docker swarm join --token SWMTKN-1-60am9y6axwot0angn1e5inxrpzrj5d6aa91gx72f8et94wztm1-7lz0dth35wywekjd1qn30jtes 192.168.1.5:2377

We use these commands to establish a Docker cluster for our service development, and add the Docker of relevant development colleagues to this cluster, completing the first step of building a cross-host network.

Establishing a cross-host network

Next, we will use docker network create command to create the Overlay network.

$ sudo docker network create --driver overlay --attachable mongodbs

When creating the Overlay network, we need to add the --attachable option so that Docker containers on different machines can use it normally.

After creating this network, we can use docker network ls on any Docker instance joined to the cluster to view the network list under it. We will find that this network definition has been synchronized to all nodes in the cluster.

$ sudo docker network ls
NETWORK ID NAME DRIVER SCOPE
## ......
y89bt74ld9l8 mongodbs overlay swarm
## ......

Next, we need to modify the Docker Compose definition so that it uses the network we have already defined instead of creating a new network.

We only need to set the network's external property to true in the network definition section of the Docker Compose configuration file, and Docker Compose will connect all the containers it creates to this project that does not belong to Docker Compose.

networks:
 mesh:
  external: true

Through this implementation, we put the entire service in a network that can use alias mapping during development, avoiding the tedious process of switching service IPs when debugging different functions. Under this structure, we only need to let the Docker we developed exit and join different clusters to immediately switch between different joint debugging projects.

2. Write a docker-compose file

Master Node

version: "3"
services: 
 master:
  image: mongo:4.1
  container_name: master
  environment:
   MONGO_INITDB_ROOT_USERNAME: root
   MONGO_INITDB_ROOT_PASSWORD: 123456
   TZ: "Asia/Shanghai"
  volumes:
   # Mount the MongoDB data directory - "/data/docker/mongodb/data/mongo:/data/db:rw"
   # Mount KeyFile
   - "/data/docker/mongodb/data/mongodb.key:/data/mongodb.key"
  ports:
   - "27018:27017"
  networks:
   - mongodbs
  command:
   # Password --auth
   #Replica set name --replSet testSet 
   --oplogSize 128
   --keyFile /data/mongodb.key
# Swarm cross-host networks networks:
 mongodbs:
  external: true

Secondary Node

version: "3"
services: 
secondary:
 image: mongo:4.1
 container_name: secondary
 environment:
  MONGO_INITDB_ROOT_USERNAME: root
  MONGO_INITDB_ROOT_PASSWORD: 123456
  TZ: "Asia/Shanghai"
 volumes:
  - "/data/docker/mongodb/data/mongo:/data/db:rw"
  - "/data/docker/mongodb/data/mongodb.key:/data/mongodb.key"
 ports:
  - "27018:27017"
 networks:
  - mongodbs
 command:
  --auth
  --replSet testSet 
  --oplogSize 128
  --keyFile /data/mongodb.key
networks:
mongodbs:
 external: true

Arbitration nodes do not need to store data. They are only used to elect a new master node when the master node fails. Therefore, they do not require passwords or port mapping operations.

version: "3"
services:
arbiter:
 image: mongo:4.1
 container_name: arbiter
 restart: always
 volumes:
  - "/data/docker/mongodb/data/mongo:/data/db:rw"
  - "/data/docker/mongodb/data/mongo_key:/mongo:rw"
 networks:
  - mongodbs
 command:
  mongod --replSet testSet --smallfiles --oplogSize 128
networks:
mongodbs:
 external: true

3. Start the container

Next, we use container orchestration to start containers in three servers respectively.

docker-compose up -d

4. Configure Replica Set

Enter the master node container

docker exec -it master mongo

Execute in the mongo shell:

> rs.initiate()
{
   "info2" : "no configuration specified. Using a default configuration for the set",
   "me" : "7abd89794aa7:27017",
   "ok" : 1
}

Continue execution:

testSet:SECONDARY> rs.add('secondary:27017')
{
   "ok" : 1,
   "$clusterTime" : {
       "clusterTime" : Timestamp(1599562800, 1),
       "signature" : {
           "hash" : BinData(0,"wrxMUIX/0bEyLgCVoQqdLvH59T0="),
           "keyId" : NumberLong("6870069879538450434")
       }
   },
   "operationTime" : Timestamp(1599562800, 1)
}

Continue to execute, where true means that this node is an arbitration node

testSet:PRIMARY> rs.add('arbiter:27017',true)
{
   "ok" : 1,
   "$clusterTime" : {
       "clusterTime" : Timestamp(1599562838, 1),
       "signature" : {
           "hash" : BinData(0,"p9ub49lLD8ij8nkxpfu2l/AvRRY="),
           "keyId" : NumberLong("6870069879538450434")
       }
   },
   "operationTime" : Timestamp(1599562838, 1)
}

View Configuration

testSet:PRIMARY> rs.conf()
{
   "_id" : "testSet",
   "version" : 3,
   "protocolVersion" : NumberLong(1),
   "writeConcernMajorityJournalDefault" : true,
   "members" : [
       {
           "_id" : 0,
           "host" : "7abd89794aa7:27017",
           "arbiterOnly" : false,
           "buildIndexes" : true,
           "hidden" : false,
           "priority" : 1,
           "tags" : {

           },
           "slaveDelay" : NumberLong(0),
           "votes" : 1
       },
       {
           "_id" : 1,
           "host" : "secondary:27017",
           "arbiterOnly" : false,
           "buildIndexes" : true,
           "hidden" : false,
           "priority" : 1,
           "tags" : {

           },
           "slaveDelay" : NumberLong(0),
           "votes" : 1
       },
       {
           "_id" : 2,
           "host" : "arbiter:27017",
           "arbiterOnly" : true,
           "buildIndexes" : true,
           "hidden" : false,
           "priority" : 0,
           "tags" : {

           },
           "slaveDelay" : NumberLong(0),
           "votes" : 1
       }
   ],
   "settings" : {
       "chainingAllowed" : true,
       "heartbeatIntervalMillis" : 2000,
       "heartbeatTimeoutSecs" : 10,
       "electionTimeoutMillis" : 10000,
       "catchUpTimeoutMillis" : -1,
       "catchUpTakeoverDelayMillis" : 30000,
       "getLastErrorModes" : {

       },
       "getLastErrorDefaults" : {
           "w" : 1,
           "wtimeout" : 0
       },
       "replicaSetId" : ObjectId("5f576426fe90ef2dd8cd2700")
   }
}

View Status

testSet:PRIMARY> rs.status()
{
   "set" : "testSet",
   "date" : ISODate("2020-09-08T11:45:12.096Z"),
   "myState" : 1,
   "term" : NumberLong(1),
   "syncingTo" : "",
   "syncSourceHost" : "",
   "syncSourceId" : -1,
   "heartbeatIntervalMillis" : NumberLong(2000),
   "optimes" : {
       "lastCommittedOpTime" : {
           "ts" : Timestamp(1599565502, 1),
           "t" : NumberLong(1)
       },
       "lastCommittedWallTime" : ISODate("2020-09-08T11:45:02.775Z"),
       "readConcernMajorityOpTime" : {
           "ts" : Timestamp(1599565502, 1),
           "t" : NumberLong(1)
       },
       "readConcernMajorityWallTime" : ISODate("2020-09-08T11:45:02.775Z"),
       "appliedOpTime" : {
           "ts" : Timestamp(1599565502, 1),
           "t" : NumberLong(1)
       },
       "durableOpTime" : {
           "ts" : Timestamp(1599565502, 1),
           "t" : NumberLong(1)
       },
       "lastAppliedWallTime" : ISODate("2020-09-08T11:45:02.775Z"),
       "lastDurableWallTime" : ISODate("2020-09-08T11:45:02.775Z")
   },
   "lastStableRecoveryTimestamp" : Timestamp(1599565492, 1),
   "lastStableCheckpointTimestamp" : Timestamp(1599565492, 1),
   "members" : [
       {
           "_id" : 0,
           "name" : "7abd89794aa7:27017",
           "ip" : "10.0.1.41",
           "health" : 1,
           "state" : 1,
           "stateStr" : "PRIMARY",
           "uptime" : 2784,
           "optime" : {
               "ts" : Timestamp(1599565502, 1),
               "t" : NumberLong(1)
           },
           "optimeDate" : ISODate("2020-09-08T11:45:02Z"),
           "syncingTo" : "",
           "syncSourceHost" : "",
           "syncSourceId" : -1,
           "infoMessage" : "",
           "electionTime" : Timestamp(1599562790, 2),
           "electionDate" : ISODate("2020-09-08T10:59:50Z"),
           "configVersion" : 3,
           "self" : true,
           "lastHeartbeatMessage" : ""
       },
       {
           "_id" : 1,
           "name" : "secondary:27017",
           "ip" : "10.0.1.233",
           "health" : 1,
           "state" : 2,
           "stateStr" : "SECONDARY",
           "uptime" : 2711,
           "optime" : {
               "ts" : Timestamp(1599565502, 1),
               "t" : NumberLong(1)
           },
           "optimeDurable" : {
               "ts" : Timestamp(1599565502, 1),
               "t" : NumberLong(1)
           },
           "optimeDate" : ISODate("2020-09-08T11:45:02Z"),
           "optimeDurableDate" : ISODate("2020-09-08T11:45:02Z"),
           "lastHeartbeat" : ISODate("2020-09-08T11:45:11.494Z"),
           "lastHeartbeatRecv" : ISODate("2020-09-08T11:45:11.475Z"),
           "pingMs" : NumberLong(0),
           "lastHeartbeatMessage" : "",
           "syncingTo" : "7abd89794aa7:27017",
           "syncSourceHost" : "7abd89794aa7:27017",
           "syncSourceId" : 0,
           "infoMessage" : "",
           "configVersion" : 3
       },
       {
           "_id" : 2,
           "name" : "arbiter:27017",
           "ip" : null,
           "health" : 0,
           "state" : 8,
           "stateStr" : "(not reachable/healthy)",
           "uptime" : 0,
           "lastHeartbeat" : ISODate("2020-09-08T11:45:10.463Z"),
           "lastHeartbeatRecv" : ISODate("1970-01-01T00:00:00Z"),
           "pingMs" : NumberLong(0),
           "lastHeartbeatMessage" : "Error connecting to arbiter:27017 :: caused by :: Could not find address for arbiter SocketException: Host not found (authoritative)",
           "syncingTo" : "",
           "syncSourceHost" : "",
           "syncSourceId" : -1,
           "infoMessage" : "",
           "configVersion" : -1
       }
   ],
   "ok" : 1,
   "$clusterTime" : {
       "clusterTime" : Timestamp(1599565502, 1),
       "signature" : {
           "hash" : BinData(0,"7/ei+8UrhlpIny9zKeWuAFpn46c="),
           "keyId" : NumberLong("6870069879538450434")
       }
   },
   "operationTime" : Timestamp(1599565502, 1)
}

5. Verify MongoDB availability

First enter the master node server to add a piece of data

docker exec -it master mongo
use admin
db.auth('root', '123456')
use test
db.test.insert({name:"muyang",age:20})

Check on the secondary node server whether the data has been synchronized.

[root@linux secondary] docker exec -it secondary mongo
testSet:SECONDARY> use admin
testSet:SECONDARY> db.auth('root', '123456')
testSet:SECONDARY> use test
testSet:SECONDARY> db.test.find()
2020-09-08T19:03:02.295+0800 E QUERY [js] uncaught exception: Error: listCollections failed: {
   "operationTime" : Timestamp(1599562972, 1),
   "ok" : 0,
   "errmsg" : "not master and slaveOk=false",
   "code" : 13435,
   "codeName" : "NotMasterNoSlaveOk",
   "$clusterTime" : {
       "clusterTime" : Timestamp(1599562972, 1),
       "signature" : {
           "hash" : BinData(0,"mhsrpGHRl7qZg2QOjyS3RbBb/Yc="),
           "keyId" : NumberLong("6870069879538450434")
       }
   }
} :
testSet:SECONDARY> rs.slaveOk()
testSet:SECONDARY> db.users.find()
{ "_id" : ObjectId("5f5764b1f909544b783696c2"), "name" : "muyang", "age" : 20 }

The following error is reported during the secondary query:

not master and slaveok=false

This is normal, because secondary is not allowed to read or write. If you must solve it, the method is as follows:

testSet:SECONDARY> rs.slaveOk()

This is the end of this article about the implementation steps of Docker to build a cluster MongoDB. For more relevant content about Docker to build a cluster MongoDB, please search for previous articles on 123WORDPRESS.COM or continue to browse the following related articles. I hope you will support 123WORDPRESS.COM in the future!

You may also be interested in:
  • How to install the latest version of MongoDB using Docker
  • Docker container deployment attempt - multi-container communication (node+mongoDB+nginx)
  • Detailed explanation of using mongodb database in docker (access in LAN)
  • Method for implementing authorized access to MongoDB based on Docker
  • How to deploy MongoDB container with Docker

<<:  Vue implements the packaging and use of components to control the number of goods

>>:  Advantages and disadvantages of MySQL indexes and guidelines for creating indexes

Recommend

HTML code that can make IE freeze

We simply need to open any text editor, copy the f...

Analysis of the difference between HTML relative path and absolute path

HTML beginners often encounter the problem of how ...

js code that associates the button with the enter key

Copy code The code is as follows: <html> &l...

Learn to deploy microservices with docker in ten minutes

Since its release in 2013, Docker has been widely...

Tips for Mixing OR and AND in SQL Statements

Today, there is such a requirement. If the logged...

Solution to the problem that the background image of a label does not display in IE8

I encountered a small problem today and struggled ...

WeChat applet custom scroll-view example code

Mini Program Custom Scroll-View Scroll Bar Withou...

Summary of ten Linux command aliases that can improve efficiency

Preface Engineers working in the Linux environmen...

Using vue3 to implement counting function component encapsulation example

Table of contents Preface 1. The significance of ...

Several solutions for CSS record text icon alignment

It is very common to see images and text displaye...

Solution to the problem of stuck when clicking Select All in Element's shuttle box with large amount of data

Table of contents Solution 1: Copy the transfer c...

MySQL error: Deadlock found when trying to get lock; try restarting transaction solution

Find the problem Recently, when I was filling in ...

How to handle images in Vue forms

question: I have a form in Vue for uploading blog...

HTML Tutorial: Collection of commonly used HTML tags (5)

These introduced HTML tags do not necessarily ful...

Solution to the garbled code problem in MySQL 5.x

MySQL is a commonly used open source database sof...