A enterprise security architect is configuring an automated cross-region database failover workflow between two active-passive data centers to maintain zero Recovery Point Objective (RPO) and minimal Recovery Time Objective (RTO) during an ungraceful outage. Arrange the operational steps of the automated failover sequence in the correct chronological order from first step executed to final step completed.
- 1Evaluate cluster quorum consensus via an independent witness site to prevent split-brain conditions.
- 2Execute node fencing (STONITH) to forcibly isolate and revoke write access from the failed primary database instance.
- 3Verify write-ahead log sequence numbers (LSN) on the standby node to validate sync replication completion.
- 4Promote the standby database instance to read-write state and initialize local database services.
- 5Update Global Server Load Balancing (GSLB) health checks and DNS pointer records to direct application traffic to the promoted site.
Answer
The correct operational sequence for automated database failover is: 1) Evaluate cluster quorum consensus via an independent witness site; 2) Execute node fencing (STONITH) to forcibly isolate the failed primary database; 3) Verify write-ahead log sequence numbers (LSN) on the standby node; 4) Promote the standby database instance to read-write state; 5) Update Global Server Load Balancing (GSLB) health checks and DNS records.
The sequence follows standard high-availability resiliency protocols: Cluster quorum must first be established to prevent split-brain, followed immediately by STONITH fencing of the legacy node. Once fenced, data consistency (LSN sync) must be verified on the target node before promoting it to primary. Finally, global network routing (GSLB/DNS) is updated to steer client traffic to the newly active environment.
Step-by-Step Solution
Key Concept
High-Availability Failover & Fencing Sequences