A database system is configured with a high availability solution where a primary server handles all database operations, and a secondary server maintains an exact, transactionally consistent copy of the primary database. In the event of a primary server failure, client applications are automatically redirected to the secondary server, which immediately takes over as the new primary without any data loss. This setup most accurately describes which type of high availability architecture?
- ALog Shipping
- BActive-Active Cluster
- CFailover Cluster Instance (FCI)
- DDatabase Mirroring (synchronous)
Show answer & explanationAnswer & explanation
Correct answer: D. Database Mirroring (synchronous)
Synchronous Database Mirroring (or similar synchronous replication in other systems like Always On Availability Groups) achieves an exact, transactionally consistent copy and automatic failover without data loss. It typically involves a principal (primary) and a mirror (secondary) server, with a witness for automatic failover. This fits the description better than a generic FCI which might use shared storage.
Why the other options are wrong
- A. Log shipping is primarily for disaster recovery and typically involves some data loss (non-zero RPO) and often manual failover, which does not meet the 'no data loss' and 'automatically redirected' requirements.
- B. Active-Active clusters distribute workload and allow multiple nodes to be primary, which doesn't fit the 'primary and secondary' model with automatic redirection to a new primary.
- C. A Failover Cluster Instance (FCI) typically relies on shared storage, where the 'instance' moves between nodes. The scenario describes maintaining an 'exact, transactionally consistent copy' which implies data replication, not just shared storage access.
Synchronous Database Mirroring
Synchronous Database Mirroring is a high-availability solution that maintains two copies of a single database, typically on different servers. Transactions are committed on both the principal (primary) and mirror (secondary) databases before being acknowledged, ensuring zero data loss and enabling automatic failover.
- Maintains two identical copies of a database.
- Transactions committed on both primary and secondary before acknowledgment.
- Provides zero data loss (RPO=0) and automatic failover.
Memory trick: Mirror, Sync, Always On: Your data's twin, never gone.