There are probably as many distributed database configurations as there are distributed databases. This chapter provides an overview of the most common problems that distributed database systems can solve and discusses the choices and trade-offs associated with each. I hope that one or more of these sample architectures corresponds to yours.
The generally accepted definition of a high-availability system is one that is operational 99.9% of the time, which translates to no more than 8 hours and 45 minutes of downtime per year. Most hardware vendors have products that are designed to ensure the high availability of servers, disk drives, and other components. In the case of servers, the recommended solution is usually a clustering technology. In some cases, high-availability systems deliver the added benefit of scalability because they require redundant computers that can share the workload.
Designing high availability into a database system, however, takes more than just buying high-availability hardware. Oracle gives you three choices for creating a high-availability system:
A hot standby database
Oracle parallel server (OPS)
Of these three, the only one that is really a distributed database is the advanced replication solution.
Oracle’s hot standby database solution can best be described as a database that is in a state of perpetual media recovery. The strategy is to create a backup ...