Prerequisites for Disaster Recovery

To use the disaster recovery feature successfully, your enterprise must meet the following requirements.

• Equivalent Hardware. Outfit the disaster recovery site with hardware equivalent to equipment at the main site.

• WAN Capacity. Link the two sites with sufficient bandwidth to replicate application data in real time, even during periods of peak volume. It is good practice to link the sites with redundant WAN lines for fault tolerance.

• Parallel Sets of FTL Servers. Run an equivalent set of FTL server processes at each site.

• Parallel Sets of Persistence Services. Configure an equivalent set of persistence services at each site. Clients connect only to services in the primary set at the main site. Standby services at the disaster recovery site replicate data, but clients do not connect to them during normal operation.

• Defined Procedures for Unplanned Failover. FTL cannot define what constitutes a disaster for your enterprise. You must have procedures for determining that the primary site is unrecoverable and that a failover to the DR site is necessary. When a DR failover has been executed, you must have procedures for ensuring that the primary site does not come back and interact with client applications in a harmful way. When the DR site is activated, any changes to application data at the original primary site cannot be recovered or replicated to the DR site.

• Defined Procedures for Client Failover. After activating the DR site, you must direct clients to the DR site, where they resume operation. As an example, you could remap DNS in your enterprise, or restart client applications with URLs that point to the DR site.