mongodb使用docker搭建replicaSet集群与变更监听

在mongodb如果需要启用变更监听功能(watch)，mongodb需要在replicaSet或者cluster方式下运行。

replicaSet和cluster从部署难度相比，replicaSet要简单许多。如果所存储的数据量规模不算太大的情况下，那么使用replicaSet方式部署mongodb是一个不错的选择。

安装环境

mongodb版本：mongodb-6.0.5

两台主机：主机1(192.168.1.11)、主机2(192.168.1.12)

docker方式mongodb集群安装

在主机1和主机2上安装好docker，并确保两台主机能正常通信

目录与key准备

在启动mongodb前，先准备好对应的目录与访问key

#在所有主机都创建用于存储mongodb数据的文件夹
mkdir -p ~/mongo-data/{data,key,backup}
#设置key文件，用于在集群机器间互相访问，各主机的key需要保持一致
cd ~/mongo-data
#在某一节点创建key
openssl rand -base64 123 > key/mongo-rs.key
sudo chown 999 key/mongo-rs.key
#不能是755, 权限太大不行. 
sudo chmod 600 key/mongo-rs.key
#将key复制到他节点
scp key/mongo-rs.key root@192.168.1.12:/root/mongo-data/key

以上操作在各主机中创建了 ~/mongo-data/{data,key,backup} 这3个目录，且mongo-rs.key的内容一致。

运行mongodb

执行下列命令，启动mongodb

sudo docker run --name mongo --network=host -p 27017:27017 -v ~/mongo-data/data:/data/db -v ~/mongo-data/backup:/data/backup -v ~/mongo-data/key:/data/key -v /etc/localtime:/etc/localtime -e MONGO_INITDB_ROOT_USERNAME=admin -e MONGO_INITDB_ROOT_PASSWORD=123456 -d mongo:6.0.5 --replSet haiyangReplset --auth --keyFile /data/key/mongo-rs.key --bind_ip_all

上面主要将27017端口映射到主机中，并设了admin的默认密码为123456。
–replSet为指定开启replicaSet，后面跟的为副本集的名称。

配置节点

进入某一节点，进行集群配置

sudo docker exec -it mongo bash
mongosh

初始化集群前先登录验证超级管理员admin

use admin
db.auth(“admin”,“123456”)

再执行以下命令进行初始化

var config={
     _id:"haiyangReplset",
     members:[
         {_id:0,host:"192.168.1.11:27017"},
         {_id:1,host:"192.168.1.12:27017"},
]};
rs.initiate(config)

执行成功后，可以看到一个节点为主节点，另一个节点为从节点

mongosh

其他相关命令

#查看副本集状态
rs.status()
#查看副本集配置
rs.conf()
#添加节点
rs.add( { host: "ip:port"} )
#删除节点
rs.remove('ip:port')

官方客户端验证

在mongodb安装好后，再用客户端连接验证一下。
官方mongodb的客户端下载地址为：https://www.mongodb.com/try/download/compass

下载完毕后，在客户端中新建连接。
在本例中，则mongodb的连接地址为:

mongodb://admin:123456@192.168.1.11:27017,192.168.1.12:27017/?authMechanism=DEFAULT&authSource=admin&replicaSet=haiyangReplset

mongo-compass

库与监控信息一目了然~

变更监听

对于mongodb操作的api在mongodb的官网有比较完备的文档，java的文档连接为：https://www.mongodb.com/docs/drivers/java/sync/v4.9/

这里试一下mongodb中一个比较强悍的功能，记录的变更监听。
用这项功能来做一些审计的场景则会非常方便。

官方链接为：
https://www.mongodb.com/docs/drivers/java/sync/v4.9/usage-examples/watch/

这里以java客户端为例写个小demo，试一下对于mongodb中集合的创建及watch功能。

package io.github.puhaiyang;

import com.google.common.collect.Lists;
import com.mongodb.client.*;
import org.apache.commons.lang3.StringUtils;
import org.bson.Document;
import org.bson.conversions.Bson;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
import java.util.Objects;
import java.util.Optional;
import java.util.Scanner;
import java.util.concurrent.CompletableFuture;

/**
 * @author puhaiyang
 * @since 2023/3/30 20:01
 * MongodbWatchTestMain
 */
public class MongodbWatchTestMain {

    public static void main(String[] args) throws Exception {
        String uri = "mongodb://admin:123456@192.168.1.11:27017,192.168.1.12:27017/?replicaSet=haiyangReplset";
        MongoClient mongoClient = MongoClients.create(uri);
        MongoDatabase mongoDatabase = mongoClient.getDatabase("my-test-db");
        String myTestCollectionName = "myTestCollection";
        //获取出collection
        MongoCollection<Document> mongoCollection = initCollection(mongoDatabase, myTestCollectionName);
        //进行watch
        CompletableFuture.runAsync(() -> {
            while (true) {
                List<Bson> pipeline = Lists.newArrayList(
                        Aggregates.match(Filters.in("ns.coll", myTestCollectionName)),
                        Aggregates.match(Filters.in("operationType", Arrays.asList("insert", "update", "replace", "delete")))
                );
                ChangeStreamIterable<Document> changeStream = mongoDatabase.watch(pipeline)
                        .fullDocument(FullDocument.UPDATE_LOOKUP)
                        .fullDocumentBeforeChange(FullDocumentBeforeChange.WHEN_AVAILABLE);

                changeStream.forEach(event -> {
                    String collectionName = Objects.requireNonNull(event.getNamespace()).getCollectionName();
                    System.out.println("--------> event:" + event.toString());
                });
            }
        });

        //数据变更测试
        {
            Thread.sleep(3_000);
            InsertOneResult insertResult = mongoCollection.insertOne(new Document("test", "sample movie document"));
            System.out.println("Success! Inserted document id: " + insertResult.getInsertedId());
            UpdateResult updateResult = mongoCollection.updateOne(new Document("test", "sample movie document"), Updates.set("field2", "sample movie document update"));
            System.out.println("Updated " + updateResult.getModifiedCount() + " document.");
            DeleteResult deleteResult = mongoCollection.deleteOne(new Document("field2", "sample movie document update"));
            System.out.println("Deleted " + deleteResult.getDeletedCount() + " document.");
        }

        new Scanner(System.in).next();
    }

    private static MongoCollection<Document> initCollection(MongoDatabase mongoDatabase, String myTestCollectionName) {
        ArrayList<Document> existsCollections = mongoDatabase.listCollections().into(new ArrayList<>());
        Optional<Document> existsCollInfoOpl = existsCollections.stream().filter(doc -> StringUtils.equals(myTestCollectionName, doc.getString("name"))).findFirst();
        existsCollInfoOpl.ifPresent(collInfo -> {
            //确保开启了changeStreamPreAndPost
            Document changeStreamPreAndPostImagesEnable = collInfo.get("options", Document.class).get("changeStreamPreAndPostImages", Document.class);
            if (changeStreamPreAndPostImagesEnable != null && !changeStreamPreAndPostImagesEnable.getBoolean("enabled")) {
                Document mod = new Document();
                mod.put("collMod", myTestCollectionName);
                mod.put("changeStreamPreAndPostImages", new Document("enabled", true));
                mongoDatabase.runCommand(mod);
            }
        });
        if (!existsCollInfoOpl.isPresent()) {
            CreateCollectionOptions collectionOptions = new CreateCollectionOptions();
            //创建collection时开启ChangeStreamPreAndPostImages
            collectionOptions.changeStreamPreAndPostImagesOptions(new ChangeStreamPreAndPostImagesOptions(true));
            mongoDatabase.createCollection(myTestCollectionName, collectionOptions);
        }
        return mongoDatabase.getCollection(myTestCollectionName);
    }
}

输出结果如下：

--------> event:ChangeStreamDocument{ operationType=insert, resumeToken={"_data": "8264255A0F000000022B022C0100296E5A10046A3E3757D6A64DF59E6D94DC56A9210446645F6964006464255A105A91F005CFB2E6D20004"}, namespace=my-test-db.myTestCollection, destinationNamespace=null, fullDocument=Document{{_id=64255a105a91f005cfb2e6d2, test=sample movie document}}, fullDocumentBeforeChange=null, documentKey={"_id": {"$oid": "64255a105a91f005cfb2e6d2"}}, clusterTime=Timestamp{value=7216272998402097154, seconds=1680169487, inc=2}, updateDescription=null, txnNumber=null, lsid=null, wallTime=BsonDateTime{value=1680169487686}}
Success! Inserted document id: BsonObjectId{value=64255a105a91f005cfb2e6d2}
Updated 1 document.
--------> event:ChangeStreamDocument{ operationType=update, resumeToken={"_data": "8264255A0F000000032B022C0100296E5A10046A3E3757D6A64DF59E6D94DC56A9210446645F6964006464255A105A91F005CFB2E6D20004"}, namespace=my-test-db.myTestCollection, destinationNamespace=null, fullDocument=Document{{_id=64255a105a91f005cfb2e6d2, test=sample movie document, field2=sample movie document update}}, fullDocumentBeforeChange=Document{{_id=64255a105a91f005cfb2e6d2, test=sample movie document}}, documentKey={"_id": {"$oid": "64255a105a91f005cfb2e6d2"}}, clusterTime=Timestamp{value=7216272998402097155, seconds=1680169487, inc=3}, updateDescription=UpdateDescription{removedFields=[], updatedFields={"field2": "sample movie document update"}, truncatedArrays=[], disambiguatedPaths=null}, txnNumber=null, lsid=null, wallTime=BsonDateTime{value=1680169487708}}
--------> event:ChangeStreamDocument{ operationType=delete, resumeToken={"_data": "8264255A0F000000042B022C0100296E5A10046A3E3757D6A64DF59E6D94DC56A9210446645F6964006464255A105A91F005CFB2E6D20004"}, namespace=my-test-db.myTestCollection, destinationNamespace=null, fullDocument=null, fullDocumentBeforeChange=Document{{_id=64255a105a91f005cfb2e6d2, test=sample movie document, field2=sample movie document update}}, documentKey={"_id": {"$oid": "64255a105a91f005cfb2e6d2"}}, clusterTime=Timestamp{value=7216272998402097156, seconds=1680169487, inc=4}, updateDescription=null, txnNumber=null, lsid=null, wallTime=BsonDateTime{value=1680169487721}}
Deleted 1 document.