MongoDB Error 6

Error code

6

Code name

HostUnreachable

Product

MongoDB

Versions

MongoDB 7.0 and 8.0 (wordings checked against the 8.0 source)

Category

Replication & Connectivity

Last reviewed

24 October 2026

What it means

Error 6 (HostUnreachable) means a MongoDB server could not reach another MongoDB process it needed: a replica set member, a shard or a config server. The text after "caused by" is the network reason, such as Connection refused.

This is mainly a server-to-server error. Replica set members send each other heartbeats, secondaries pull the oplog from a sync source, and mongos talks to shards and config servers. When one of those connections fails, MongoDB 8.0's internal connection pool reports the failure as HostUnreachable and keeps the original reason. The message names the host as configured and, in brackets, the address it resolved to, then the cause: Connection refused (nothing listening on that port), a time-out, or a closed or reset connection. The cause text after the last caused by comes from the operating system or network layer, so its wording can vary by platform.

Applications usually do not see code 6 directly. A driver that cannot reach a server reports its own error, typically a server selection timeout; code 6 appears in rs.status() (in a member's lastHeartbeatMessage), in the server log, and in errors that mongos passes back when a shard cannot be reached. error_codes.yml puts it in the NetworkError and RetriableError categories.

The documentation's replica set troubleshooting guide is the reference here: every member must be able to connect to every other member, in both directions, and a mongod bound only to localhost does not accept connections from other hosts, including other members.

Common causes

• The target mongod or mongos is stopped, crashed or still starting, so the port refuses connections.
• A firewall, security group or network ACL blocks the port between the two hosts, often in only one direction.
• The target listens only on localhost (net.bindIp) or on another port than the one in the replica set configuration.
• The host name in the replica set configuration resolves to a wrong or unreachable address from this member.
• A network partition, or connections closed by a load balancer or proxy between members.

Diagnose it

Read-only These mongosh commands only read data, indexes, statistics and server state (find, aggregate without $out or $merge, getIndexes, stats, currentOp, rs.status). They do not change data, users or settings, so they are safe to run first.

rs.status().members.map(m => ({ name: m.name, health: m.health, state: m.stateStr, msg: m.lastHeartbeatMessage }))

rs.conf().members.map(m => m.host)

mongosh --host m2.example.net --port 27017 --eval "db.hello().me"

// on the target member
db.adminCommand({ getCmdLineOpts: 1 }).parsed.net

How to fix it

Each fix is labelled with what it changes. Try read-only checks and the least invasive fix first, and test changes outside production.

# Linux (systemd)
sudo systemctl start mongod

# Windows (PowerShell, default service name)
Start-Service MongoDB

cfg = rs.conf()
cfg.members[1].host = "m2.example.net:27017"
rs.reconfig(cfg)

Examples

A member that is down, seen in rs.status()

Not reproduced here: no MongoDB server is installed for this site. The field names are from the replSetGetStatus documentation; the heartbeat message follows the format built by makeConnectError() in MongoDB 8.0's asio_transport_layer.cpp, which the connection pool reports as HostUnreachable. The state text and host names are illustrative.

rs.status().members[1]

{
  name: 'm2.example.net:27017',
  health: 0,
  stateStr: '(not reachable/healthy)',
  lastHeartbeatMessage: 'Error connecting to m2.example.net:27017 (10.0.0.12:27017) :: caused by :: Connection refused',
  ...
}

Choosing a sync source that cannot be reached

Not reproduced here: no MongoDB server is installed for this site. The text is built for the replSetSyncFrom command in MongoDB 8.0's topology_coordinator.cpp with code HostUnreachable when the requested member is not reachable from this one; the host name is illustrative.

rs.syncFrom("m3.example.net:27017")

MongoServerError: I cannot reach the requested member: m3.example.net:27017

How to prevent it

Use stable, resolvable host names for members, monitor member health and heartbeat messages, and test connectivity in both directions whenever firewall rules or networks change.

Keep members on the same port and bind configuration, and document which hosts must reach which, including mongos to every shard and config server.

Common questions

Learn more on CodeWithSQL

• Install MongoDB

Official documentation

• Troubleshoot replica sets (test connections between members)
• replSetGetStatus (lastHeartbeatMessage)
• Security checklist and IP binding
• MongoDB error codes (6 HostUnreachable)

Last technically reviewed 24 October 2026; examples and diagnostics run on MongoDB 8.0 documentation and source (error_codes.yml r8.0.0); not run locally. Codes and code names are checked against MongoDB's error_codes.yml (8.0); no MongoDB server was available, so the examples were not run locally.