Overview of replication
Replication is a data synchronization mechanism that ensures that updates made to a database are automatically duplicated to other servers.
Replication improves data availability when unforeseen or planned outages occur and improves search performance by allowing client requests to be distributed across multiple servers.
By default, all PingDirectory server instances participating in replication are writable so that LDAP clients can perform updates at any of these server instances. These updates are automatically propagated to the other servers in the background in the same order as the updates were entered. The replication process flow is designed to immediately propagate changes to the other replicas in the topology with little or no latency.
The following figure demonstrates the basic flow of replication.
The benefits of replication can be summarized as follows:
- High-Availability
-
Because the data is fully replicated on all other servers in the topology, replication allows participating servers to process all types of client requests. This mitigates any availability issues when a particular server is down because of a planned maintenance or unplanned outage. Servers that are temporarily unavailable receive updates when they become available again.
- Improved Search Performance
-
Search requests can be directed to any PingDirectory server participating in replication, which improves search performance over systems that only access single servers.
|
Replication does not improve write throughput because updates need to be applied at all servers. |
- WAN Friendly Data Synchronization
-
The built-in compression feature in the replication protocol allows efficient propagation of updates over WAN links.