使用 Docker 部署 Mongodb 集群后项目中的连接问题

问题

CentOS 7 + docker + mongodb 4.4.4 部署了一个 mongodb 集群，然后在项目(Java-springboot)中去连接 mongodb 进行数据的读写，出现如下提示：

2021-05-12 13:42:33.003  INFO 17864 --- [io-8084-exec-10] org.mongodb.driver.cluster               : Cluster created with settings {hosts=[121.*.*.56:27017], mode=MULTIPLE, requiredClusterType=UNKNOWN, serverSelectionTimeout='30000 ms', maxWaitQueueSize=1500000}
2021-05-12 13:42:33.003  INFO 17864 --- [io-8084-exec-10] org.mongodb.driver.cluster               : Adding discovered server 121.*.*.56:27017 to client view of cluster
2021-05-12 13:42:33.006  INFO 17864 --- [io-8084-exec-10] org.mongodb.driver.cluster               : No server chosen by WritableServerSelector from cluster description ClusterDescription{type=UNKNOWN, connectionMode=MULTIPLE, serverDescriptions=[ServerDescription{address=121.36.16.56:27017, type=UNKNOWN, state=CONNECTING}]}. Waiting for 30000 ms before timing out
2021-05-12 13:42:33.108  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.connection            : Opened connection [connectionId{localValue:985, serverValue:377}] to 121.*.*.56:27017
2021-05-12 13:42:33.133  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.cluster               : Monitor thread successfully connected to server with description ServerDescription{address=121.*.*.56:27017, type=REPLICA_SET_PRIMARY, state=CONNECTED, ok=true, version=ServerVersion{versionList=[4, 4, 4]}, minWireVersion=0, maxWireVersion=9, maxDocumentSize=16777216, roundTripTimeNanos=25262872, setName='mongoreplset', canonicalAddress=172.17.0.4:27017, hosts=[172.17.0.4:27017, 121.*.*.103:27017, 172.17.0.6:27017], passives=[], arbiters=[], primary='172.17.0.4:27017', tagSet=TagSet{[]}, electionId=7fffffff0000000000000024, setVersion=2, lastWriteDate=Wed May 12 13:42:42 CST 2021, lastUpdateTimeNanos=533944664886564}
2021-05-12 13:42:33.134  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.cluster               : Discovered cluster type of REPLICA_SET
2021-05-12 13:42:33.134  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.cluster               : Adding discovered server 172.17.0.4:27017 to client view of cluster
2021-05-12 13:42:33.135  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.cluster               : Adding discovered server 121.*.*.103:27017 to client view of cluster
2021-05-12 13:42:33.136  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.cluster               : Adding discovered server 172.17.0.6:27017 to client view of cluster
2021-05-12 13:42:33.137  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.cluster               : Server 121.*.*.56:27017 is no longer a member of the replica set.  Removing from client view of cluster.
2021-05-12 13:42:33.137  INFO 17864 --- [.*.*.56:27017] org.mongodb.driver.cluster               : Canonical address 172.17.0.4:27017 does not match server address.  Removing 121.36.16.56:27017 from client view of cluster
2021-05-12 13:42:34.212  INFO 17864 --- [*.*.103:27017] org.mongodb.driver.cluster               : Exception in monitor thread while connecting to server 121.*.*.103:27017
​

进入 mongodb 容器中，查看当前 rs.conf()

mongoreplset:PRIMARY> rs.conf()
{
        "_id" : "mongoreplset",
        "version" : 2,
        "term" : 36,
        "protocolVersion" : NumberLong(1),
        "writeConcernMajorityJournalDefault" : true,
        "members" : [
                {
                        "_id" : 0,
                        "host" : "172.17.0.4:27017",
                        "arbiterOnly" : false,
                        "buildIndexes" : true,
                        "hidden" : false,
                        "priority" : 1,
                        "tags" : {
                      
                        },
                        "slaveDelay" : NumberLong(0),
                        "votes" : 1
                },
                {
                        "_id" : 1,
                        "host" : "172.17.0.6:27017",
                        "arbiterOnly" : false,
                        "buildIndexes" : true,
                        "hidden" : false,
                        "priority" : 1,
                        "tags" : {
                      
                        },
                        "slaveDelay" : NumberLong(0),
                        "votes" : 1
                },
		...
​

可以看到现在的复制集中使用的是非外网 IP 配置的，此时，在项目中连接时就会出现上面的连接不通问题。

后来重新部署，然后将其中的 host 换为服务器 IP:port 后，就可以正常连接读写数据了。

问题：

是否只能如此配置外网 ip?

如果配置宿主机内的 ip ，应该如何配置才能不报错，是不是需要配置 宿主机上的 hosts 映射到内网 ip ?