To interface with the replication topology, the Directory Server provides a command-line utility, dsreplication, that must be used to manage and monitor replication.
The cluster name for an instance should be set to a unique name during install. The cluster-name controls a handle of configuration settings that are used cluster-wide; any changes to the configuration will always get pushed out to other members of the topology. If the cluster-name is set to a unique name for each server, any change to the configuration via the dsconfig command or the console needs to be done on every server in the topology. In a devops model, it is recommended that the cluster name be unique so that everything can be managed with server profiles. If you are doing an on-premise install, you may want to use a common cluster-name across the topology.
- Set up the servers. This is the basic installation steps to set up a Directory Server instance.
- Import or restore data to one server. After setting up the servers, at least one server should have the target data loaded through import-ldif or restore.
- Enable replication between the servers. Using the dsreplication tool, enable replication for each server to be included in the replication topology. The dsreplication enable subcommand should be run N - 1 times for a topology of N servers. See Command Line Interface for more information.
- Initialize data from source server to every server in the topology. Run the dsreplication initialize subcommand for every target server that needs a copy of the data from the source server.
- Verifying the replication topology. Administrators can check the replication status after configuring the topology using the dsreplication status tool.