Rejoin Carbonio Mesh Cluster

When a member of a Carbonio Mesh cluster stays offline for a longer time than the value of server_rejoin_age_max, it will be unable to rejoin the cluster cluster and the following error appears in syslog:

refusing to rejoin cluster because server has been offline for more than the configured server_rejoin_age_max

Hint

You can use the command journalctl -u service-discover.service -f to check the log and see new log message as they are produced, in real time.

A typical scenario for this error is when the Node, on which the member is installed, is restored from an old snapshot.

A viable solution is quite easy and requires a few commands from the CLI.

First, delete the following file.

# rm /var/lib/service-discover/data/server_metadata.json

Then restart the service-discover daemon

# systemctl restart service-discovery

If the command is successful, this message appears in syslog:

Join cluster completed.

To make sure that the Carbonio Mesh agent is synchronised with the other members, issue the following commands.

# consul members

# consul catalog services