Cannot connect to replica set via mongo compass

Thanachai_Namphairoj · 2022-05-27T03:21:05.282Z

The first one is the one that I cannot access and error is as follows

image928×566 36.4 KB

This one it works just fine

image938×473 29.8 KB

image1866×882 44.5 KB

steevej · 2022-05-27T13:10:31.074Z

I think you have an issue with your replica set configuration.

I suspect that your replica set configuration uses host names rather than the IP addresses you use to connect, and that some of those are not resolved correctly by your DNS.

Using a command line terminal share the output of the following.

ping masternode

Then connect with mongosh to a single node, 192.168.20.1:27017 for example and share the output of the command:

rs.status()

Thanachai_Namphairoj · 2022-05-30T05:24:08.702Z

Here is the result of rs.status()

image1266×522 31.7 KB

This is the result of ping masternode

image1023×446 11.4 KB

Ramachandra_Tummala · 2022-05-30T07:26:15.587Z

If this a replica set it should show all 3 members info but showing as standalone
Have you run rs.initiate() and added other 2 nodes?
Please show rs.conf() output

Thanachai_Namphairoj · 2022-05-30T07:34:10.993Z

{
set: ‘rs0’,
date: ISODate(“2022-05-30T07:31:56.942Z”),
myState: 1,
term: Long(“47”),
syncSourceHost: ‘’,
syncSourceId: -1,
heartbeatIntervalMillis: Long(“2000”),
majorityVoteCount: 2,
writeMajorityCount: 2,
votingMembersCount: 3,
writableVotingMembersCount: 3,
optimes: {
lastCommittedOpTime: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
lastCommittedWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
readConcernMajorityOpTime: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
appliedOpTime: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
durableOpTime: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
lastAppliedWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
lastDurableWallTime: ISODate(“2022-05-30T07:31:48.203Z”)
},
lastStableRecoveryTimestamp: Timestamp({ t: 1653895860, i: 1 }),
electionCandidateMetrics: {
lastElectionReason: ‘stepUpRequestSkipDryRun’,
lastElectionDate: ISODate(“2022-05-25T02:51:46.241Z”),
electionTerm: Long(“47”),
lastCommittedOpTimeAtElection: { ts: Timestamp({ t: 1653447105, i: 1 }), t: Long(“46”) },
lastSeenOpTimeAtElection: { ts: Timestamp({ t: 1653447105, i: 1 }), t: Long(“46”) },
numVotesNeeded: 2,
priorityAtElection: 1,
electionTimeoutMillis: Long(“10000”),
priorPrimaryMemberId: 1,
numCatchUpOps: Long(“0”),
newTermStartDate: ISODate(“2022-05-25T02:51:46.260Z”),
wMajorityWriteAvailabilityDate: ISODate(“2022-05-25T02:51:47.302Z”)
},
members: [
{
_id: 0,
name: ‘masternode:27017’,
health: 1,
state: 1,
stateStr: ‘PRIMARY’,
uptime: 607291,
optime: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
optimeDate: ISODate(“2022-05-30T07:31:48.000Z”),
lastAppliedWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
lastDurableWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
syncSourceHost: ‘’,
syncSourceId: -1,
infoMessage: ‘’,
electionTime: Timestamp({ t: 1653447106, i: 1 }),
electionDate: ISODate(“2022-05-25T02:51:46.000Z”),
configVersion: 1,
configTerm: 47,
self: true,
lastHeartbeatMessage: ‘’
},
{
_id: 1,
name: ‘client1:27017’,
health: 1,
state: 2,
stateStr: ‘SECONDARY’,
uptime: 448792,
optime: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
optimeDurable: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
optimeDate: ISODate(“2022-05-30T07:31:48.000Z”),
optimeDurableDate: ISODate(“2022-05-30T07:31:48.000Z”),
lastAppliedWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
lastDurableWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
lastHeartbeat: ISODate(“2022-05-30T07:31:55.333Z”),
lastHeartbeatRecv: ISODate(“2022-05-30T07:31:55.333Z”),
pingMs: Long(“0”),
lastHeartbeatMessage: ‘’,
syncSourceHost: ‘masternode:27017’,
syncSourceId: 0,
infoMessage: ‘’,
configVersion: 1,
configTerm: 47
},
{
_id: 2,
name: ‘client2:27017’,
health: 1,
state: 2,
stateStr: ‘SECONDARY’,
uptime: 607285,
optime: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
optimeDurable: { ts: Timestamp({ t: 1653895908, i: 1 }), t: Long(“47”) },
optimeDate: ISODate(“2022-05-30T07:31:48.000Z”),
optimeDurableDate: ISODate(“2022-05-30T07:31:48.000Z”),
lastAppliedWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
lastDurableWallTime: ISODate(“2022-05-30T07:31:48.203Z”),
lastHeartbeat: ISODate(“2022-05-30T07:31:55.295Z”),
lastHeartbeatRecv: ISODate(“2022-05-30T07:31:54.984Z”),
pingMs: Long(“0”),
lastHeartbeatMessage: ‘’,
syncSourceHost: ‘masternode:27017’,
syncSourceId: 0,
infoMessage: ‘’,
configVersion: 1,
configTerm: 47
}
],
ok: 1,
‘$clusterTime’: {
clusterTime: Timestamp({ t: 1653895908, i: 1 }),
signature: {
hash: Binary(Buffer.from(“ca77f671a7f355a16649c47ff0d4f500f38d0e0a”, “hex”), 0),
keyId: Long(“7097159497856057348”)
}
},
operationTime: Timestamp({ t: 1653895908, i: 1 })
}

Here is my rs.status()

Thanachai_Namphairoj · 2022-05-30T07:35:00.643Z

{
_id: ‘rs0’,
version: 1,
term: 47,
members: [
{
_id: 0,
host: ‘masternode:27017’,
arbiterOnly: false,
buildIndexes: true,
hidden: false,
priority: 1,
tags: {},
secondaryDelaySecs: Long(“0”),
votes: 1
},
{
_id: 1,
host: ‘client1:27017’,
arbiterOnly: false,
buildIndexes: true,
hidden: false,
priority: 1,
tags: {},
secondaryDelaySecs: Long(“0”),
votes: 1
},
{
_id: 2,
host: ‘client2:27017’,
arbiterOnly: false,
buildIndexes: true,
hidden: false,
priority: 1,
tags: {},
secondaryDelaySecs: Long(“0”),
votes: 1
}
],
protocolVersion: Long(“1”),
writeConcernMajorityJournalDefault: true,
settings: {
chainingAllowed: true,
heartbeatIntervalMillis: 2000,
heartbeatTimeoutSecs: 10,
electionTimeoutMillis: 10000,
catchUpTimeoutMillis: -1,
catchUpTakeoverDelayMillis: 30000,
getLastErrorModes: {},
getLastErrorDefaults: { w: 1, wtimeout: 0 },
replicaSetId: ObjectId(“627e2cebd23c7aae01154b0b”)
}
}>
This is rs.config()

Ramachandra_Tummala · 2022-05-30T07:52:24.736Z

Can you connect by hostname instead IP?
Are other 2 nodes pingable and resolving to the ips you are using
Also output of cat /etc/hosts

Thanachai_Namphairoj · 2022-05-30T08:00:08.331Z

Here is the cat of /etc/hosts

image692×400 19.8 KB

I had tried using the hostname instead of IPs and it resulted in the same error.

Yes, every node can ping to each other using ping hostname and the ip is all correct.

Ramachandra_Tummala · 2022-05-30T08:26:07.128Z

Looks ok
Did you try to connect to your replica set using hostnames in your connect string?

Thanachai_Namphairoj · 2022-05-30T08:29:07.970Z

Yes I did. It gives me the same error

image915×835 57.5 KB

steevej · 2022-05-30T13:19:01.196Z

ENOTFOUND means that the host name masternode cannot be found.

In one of your previous post, you shown that you can ping masternode and the other 2 hosts of your replica set.

The only conclusion I can think of, is that you are not running Compass from the same machine as the one you used to run the ping commands.

The host names of your replica set must be DNS resolvable from all machines you are using to access the replica set. They should all resolve to IP addresses that are routed from all machines you are using to access the replica set.

Thanachai_Namphairoj · 2022-05-31T07:43:25.703Z

That might be the case because I run the replicaset on 3 separate VMs. Are there any ways for me to allow compass on my pc to be able to access to the set? I tried using ufw allow from 0.0.0.0/0 on every single VM but it still doesn’t work

steevej · 2022-05-31T13:14:21.082Z

Thanachai_Namphairoj:

compass on my pc to be able to access to the set

Sure there is. But I can only rephrase what I wrote in my previous post. The host names use in the replica set must be know by you PC and your PC must be able to route traffic to the corresponding IP addresses.

Networking, including host and domain name resolution and routing is complex. I recommend using Atlas. You would be up and running a replica set in no time.

devops_learning · 2022-11-03T16:27:20.181Z

Thank you steevej .even iam also facing same issue.

rliu · 2023-02-02T01:20:21.044Z

I have the same issue that running replica set within local docker containers while failing to connect with my local Compass outside of the docker containers.

I failed to use Atlas because my company blocked 27017 port for security reasons. To request the other team to whitelist 27017 for Atlas usage is not a option in my situation.

Since Mongo provided a detailed procedure for deploy Mongo Cluster in docker, I would hope Mongo can also provides a detailed procedure that helps out how local Compass (outside of container) connect replica sets within docker container.

steevej · 2023-02-02T21:15:00.928Z

rliu:

I would hope Mongo can also provides a detailed procedure that helps out how local Compass (outside of container) connect replica sets within docker container.

They did not specify how to connect because there is nothing special to do. The information is already there.

From the link you supplied you created a replica set named myReplicaSet which has 3 nodes, the nodes are named mongo1, mongo2 and mongo3, and they all listen on port 27017. The URI to connect to a replica set is of the form:

mongodb://[username:password@]host1[:port1][,...hostN[:portN]][/[defaultauthdb][?options]]

Just have to replace host1, …, hostN with your hosts. You may leave the port number since you are using the default one.

Running all instances of your replica set within the same physical host is risky. Because if you loose your machine you loose your data.

rliu · 2023-02-02T22:02:49.936Z

I appreciate the connection string page you shared here.

However the same issue repeated as this post-owner posted here. The views of this post has been at 4.3k in this forum, one of the most review posts here, and the same questions repeated over other technical forum too so that seems like a lot of people like me has the same issue even though we followed that connection string format.

I had been read many other posts all over internet like yours, unfortunately none of them works. That’s why I am still hoping Mongo can provide a procedure instruction.

I am doing POC for now so creating cluster inside one local machine is not my current concern for now. But thank you for the help though.

steevej · 2023-02-02T23:20:11.905Z

rliu:

none of them works

If the following connection string does not work please provide the error you are getting.

mongodb://mongo1,mongo2,mongo3/?replicaSet=myReplicaSet

What base OS are you using?

rliu · 2023-02-02T23:53:56.032Z

I appreciate your patient. Still doesn’t work

My OS is:
|Edition|Windows 10 Pro|

So far I had failed trying on the connection strings:

mongodb://localhost:27017,localhost:27018,localhost:27019/?replicaSet=myReplicaSet&tls=false

mongodb://mongo1:27017,mongo2:27018,mongo3:27019/?replicaSet=myReplicaSet&tls=false

mongodb://mongo1:27017,mongo2:27018,mongo3:27019/?replicaSet=myReplicaSet&readPreference=primary&appname=MongoDB%20Compass&directConnection=false&ssl=false

mongodb://172.20.0.1:27017,172.20.0.2:27018,172.20.0.3:27019/?replicaSet=myReplicaSet&readPreference=primary&directConnection=false&ssl=false

The above docker container IP addresses acquired by command

docker inspect container_id (9f64ba814343)

And more or less the /options here and there where I prefer not to exhaust all of my trials. The Compass always returns me error:

getaddrinfo ENOTFOUND mongo1/2/3

Note: I followed the Mongo instruction so we have common ground here at least.

steevej · 2023-02-03T02:54:08.102Z

rliu:

mongodb://localhost:27017,localhost:27018,localhost:27019/?replicaSet=myReplicaSet&tls=false

This is not what I proposed. The replica set is not configured as as localhost. So you are specifying the wrong host. From the link you supplied the replica set is:

docker exec -it mongo1 mongosh --eval "rs.initiate({
 _id: \"myReplicaSet\",
 members: [
   {_id: 0, host: \"mongo1\"},
   {_id: 1, host: \"mongo2\"},
   {_id: 2, host: \"mongo3\"}
 ]
})"

rliu:

mongodb://mongo1:27017,mongo2:27018,mongo3:27019/?replicaSet=myReplicaSet&tls=false

This is again not what I proposed. I mentioned

steevej:

the nodes are named mongo1, mongo2 and mongo3, and they all listen on port 27017

and

steevej:

You may leave the port number since you are using the default one.

So you are mentioning the wrong ports.

rliu:

mongodb://mongo1:27017,mongo2:27018,mongo3:27019/?replicaSet=myReplicaSet&readPreference=primary&appname=MongoDB%20Compass&directConnection=false&ssl=false

As already mentioned you are specifying the wrong ports.

rliu:

mongodb://172.20.0.1:27017,172.20.0.2:27018,172.20.0.3:27019/?replicaSet=myReplicaSet&readPreference=primary&directConnection=false&ssl=false

I have mentioned that your replica set, if configured like the link you shared is configured with

steevej:

the nodes are named mongo1, mongo2 and mongo3

So with 127.0.0.1 you are mentioning the wrong hosts.

The conclusion is that you have tried many things. The one thing you have not tried is the one I have mentioned:

steevej:

If the following connection string does not work please provide the error you are getting.
mongodb://mongo1,mongo2,mongo3/?replicaSet=myReplicaSet

Since you have tried many things except the one I shared I am not too sure that you

rliu:

followed the Mongo instruction

because you do not follow the only suggestion I offer. Please share the output of

docker ps 
and
docker exec -it mongo1 mongosh --eval "rs.status()"

Hopefully, they match what is documented in the link you shared.

Share the screenshot of the connection string that generated the Compass error

rliu:

getaddrinfo ENOTFOUND mongo1/2/3