Mongo Local ReplicaSet setup using Docker

Richu_John · October 14, 2024, 12:01am

HI,

I am following the documentation at Deploying A MongoDB Cluster With Docker | MongoDB

in a mac system running nerdctl as command line client to connect to containerd runtime.

I am able to complete the setup properly and connect to the database from compass.

But when I try to set the read preference as secondary, it just times out.

rjohn@FIVE9-DYFPDQ0WXH ~ % lima nerdctl exec -it mongo1 mongosh --eval "rs.status()"                                                     
{
  set: 'myReplicaSet',
  date: ISODate('2024-10-11T03:45:07.595Z'),
  myState: 1,
  term: Long('1'),
  syncSourceHost: '',
  syncSourceId: -1,
  heartbeatIntervalMillis: Long('2000'),
  majorityVoteCount: 2,
  writeMajorityCount: 2,
  votingMembersCount: 3,
  writableVotingMembersCount: 3,
  optimes: {
    lastCommittedOpTime: { ts: Timestamp({ t: 1728618307, i: 1 }), t: Long('1') },
    lastCommittedWallTime: ISODate('2024-10-11T03:45:07.056Z'),
    readConcernMajorityOpTime: { ts: Timestamp({ t: 1728618307, i: 1 }), t: Long('1') },
    appliedOpTime: { ts: Timestamp({ t: 1728618307, i: 1 }), t: Long('1') },
    durableOpTime: { ts: Timestamp({ t: 1728618307, i: 1 }), t: Long('1') },
    lastAppliedWallTime: ISODate('2024-10-11T03:45:07.056Z'),
    lastDurableWallTime: ISODate('2024-10-11T03:45:07.056Z')
  },
  lastStableRecoveryTimestamp: Timestamp({ t: 1728618287, i: 1 }),
  electionCandidateMetrics: {
    lastElectionReason: 'electionTimeout',
    lastElectionDate: ISODate('2024-10-11T03:30:06.905Z'),
    electionTerm: Long('1'),
    lastCommittedOpTimeAtElection: { ts: Timestamp({ t: 1728617396, i: 1 }), t: Long('-1') },
    lastSeenOpTimeAtElection: { ts: Timestamp({ t: 1728617396, i: 1 }), t: Long('-1') },
    numVotesNeeded: 2,
    priorityAtElection: 1,
    electionTimeoutMillis: Long('10000'),
    numCatchUpOps: Long('0'),
    newTermStartDate: ISODate('2024-10-11T03:30:06.954Z'),
    wMajorityWriteAvailabilityDate: ISODate('2024-10-11T03:30:08.126Z')
  },
  members: [
    {
      _id: 0,
      name: 'mongo1:27017',
      health: 1,
      state: 1,
      stateStr: 'PRIMARY',
      uptime: 947,
      optime: { ts: Timestamp({ t: 1728618307, i: 1 }), t: Long('1') },
      optimeDate: ISODate('2024-10-11T03:45:07.000Z'),
      lastAppliedWallTime: ISODate('2024-10-11T03:45:07.056Z'),
      lastDurableWallTime: ISODate('2024-10-11T03:45:07.056Z'),
      syncSourceHost: '',
      syncSourceId: -1,
      infoMessage: '',
      electionTime: Timestamp({ t: 1728617406, i: 1 }),
      electionDate: ISODate('2024-10-11T03:30:06.000Z'),
      configVersion: 1,
      configTerm: 1,
      self: true,
      lastHeartbeatMessage: ''
    },
    {
      _id: 1,
      name: 'mongo2:27017',
      health: 1,
      state: 2,
      stateStr: 'SECONDARY',
      uptime: 911,
      optime: { ts: Timestamp({ t: 1728618297, i: 1 }), t: Long('1') },
      optimeDurable: { ts: Timestamp({ t: 1728618297, i: 1 }), t: Long('1') },
      optimeDate: ISODate('2024-10-11T03:44:57.000Z'),
      optimeDurableDate: ISODate('2024-10-11T03:44:57.000Z'),
      lastAppliedWallTime: ISODate('2024-10-11T03:45:07.056Z'),
      lastDurableWallTime: ISODate('2024-10-11T03:45:07.056Z'),
      lastHeartbeat: ISODate('2024-10-11T03:45:05.956Z'),
      lastHeartbeatRecv: ISODate('2024-10-11T03:45:07.189Z'),
      pingMs: Long('0'),
      lastHeartbeatMessage: '',
      syncSourceHost: 'mongo1:27017',
      syncSourceId: 0,
      infoMessage: '',
      configVersion: 1,
      configTerm: 1
    },
    {
      _id: 2,
      name: 'mongo3:27017',
      health: 1,
      state: 2,
      stateStr: 'SECONDARY',
      uptime: 911,
      optime: { ts: Timestamp({ t: 1728618297, i: 1 }), t: Long('1') },
      optimeDurable: { ts: Timestamp({ t: 1728618297, i: 1 }), t: Long('1') },
      optimeDate: ISODate('2024-10-11T03:44:57.000Z'),
      optimeDurableDate: ISODate('2024-10-11T03:44:57.000Z'),
      lastAppliedWallTime: ISODate('2024-10-11T03:45:07.056Z'),
      lastDurableWallTime: ISODate('2024-10-11T03:45:07.056Z'),
      lastHeartbeat: ISODate('2024-10-11T03:45:05.957Z'),
      lastHeartbeatRecv: ISODate('2024-10-11T03:45:07.189Z'),
      pingMs: Long('0'),
      lastHeartbeatMessage: '',
      syncSourceHost: 'mongo1:27017',
      syncSourceId: 0,
      infoMessage: '',
      configVersion: 1,
      configTerm: 1
    }
  ],
  ok: 1,
  '$clusterTime': {
    clusterTime: Timestamp({ t: 1728618307, i: 1 }),
    signature: {
      hash: Binary.createFromBase64('AAAAAAAAAAAAAAAAAAAAAAAAAAA=', 0),
      keyId: Long('0')
    }
  },
  operationTime: Timestamp({ t: 1728618307, i: 1 })
}

when I try setReadPref(“secondary”), it just times out.
Could you please tell me where I am going wrong.

Billy_Bui · October 14, 2024, 3:03am

I am not familiar with Mac so I will give you some pointers to debug the situation.

Check your connection to each member of the replica set. This is because when connecting to replicaSet mode, the client will reconnect to each replica set member using the host in the config; in this case are mongo1, mongo2 and mongo3

Check if you can ping to mongo1, mongo2, mongo3 from your host machine, or from any machine that you want to run MongoDB Compass to connect to your cluster
Then check your connection to each member of the replica set by using Direct connection mode to connect to mongo1:27017, mongo2:27017, mongo3:27017

If any of above actions failed, you are having connection trouble to your cluster. In replica set mode, the initial host:port 127.0.0.1:27017 only be used for getting the replica set config, client then reconnect to each member of your cluster based on the config.

If you use lima nerdctl exec -it mongo1 mongosh to access to your cluster, I think you can run setReadPref command without issue, that because the mongosh is run inside the mongo container, which in the network that can resolve mongo1, mongo2 & mongo3.

I will suggest these actions:

Change the listening port so that each member don’t listening to the same port, for example, run these comand instead to spawn mongo containers

docker run -d --rm -p 27017:27017 --name mongo1 --network mongoCluster mongo:7 mongod --replSet myReplicaSet --bind_ip localhost,mongo1 --port 27017
docker run -d --rm -p 27018:27018 --name mongo2 --network mongoCluster mongo:7 mongod --replSet myReplicaSet --bind_ip localhost,mongo2 --port 27018
docker run -d --rm -p 27019:27019 --name mongo3 --network mongoCluster mongo:7 mongod --replSet myReplicaSet --bind_ip localhost,mongo3 --port 27019

Initiate your replica set using your host ip and above ports, for example, if your host that run containerd have ip 192.168.100.1 then initate your replica set using that ip address:

lima nerdctl exec -it mongo1 mongosh --port 27017

rs.initiate({
  _id: "myReplicaSet",
  members: [
    { _id: 0, host: "192.168.100.1:27017" },
    { _id: 1, host: "192.168.100.1:27018" },
    { _id: 2, host: "192.168.100.1:27019" },
  ]
})

Then check again if you can connect to your cluster using compass.

Billy_Bui · October 14, 2024, 3:31am

I think these 2 images can help you visualize the problem better.