Administration Guide › Set Up Distribution and Routing › DSAs with the Same Prefix › Failover and Failback › Failover between Sites

Failover between Sites

To improve performance, we recommend that each region have a router that directs queries to a local data DSA. The local data DSAs should be synchronized with each other.

If a local data DSA fails, each router directs queries to the remote data DSA. This maintains the availability and reliability of the directory.

Example: Failover between Two Sites

The following diagram shows an example directory backbone that is distributed across two sites, New York and Montreal.

The New York router cannot load-share to the Customers DSAs in Montreal because they are in a different site from the router.

However, if both of the New York Customers DSAs are unavailable, the router can failover to the Customers DSA in Montreal. Although the performance is lower until the New York DSAs are running again, the service is still available.

Note: Failover consistency is guaranteed within a region, but not between regions.