Examkin

Soru

Zorluk: ZorDisaster Recovery and Business Continuity Execution

An enterprise energy utility company operates a mission-critical smart grid monitoring platform on Google Cloud. The primary infrastructure runs in europe-west3 (Frankfurt), featuring stateless microservices on a Compute Engine Managed Instance Group (MIG) and transactional telemetry data stored in Cloud SQL for PostgreSQL. The company requires a Disaster Recovery (DR) execution plan to fail over to europe-west1 (Belgium). The architecture must guarantee a Recovery Point Objective (RPO) under 1 minute and a Recovery Time Objective (RTO) under 15 minutes, while strictly minimizing ongoing secondary region infrastructure costs during normal operational state. Which disaster recovery strategy should the Cloud Architect implement?

  1. Maintain a cross-region read replica of Cloud SQL in europe-west1, alongside a minimal Warm Standby MIG scaled to one instance using pre-validated instance templates. Upon DR declaration, promote the read replica to standalone primary, scale out the MIG to baseline capacity, and update Cloud DNS routing.Cevap
  2. B
    Implement a Cold Backup pattern by scheduling automated Cloud Storage exports of the PostgreSQL database and exporting compute VM custom images to europe-west1. During a outage, restore the database from storage objects and instantiate new MIG compute resources from the custom images.
  3. C
    Provision a Pilot Light architecture keeping Cloud SQL cross-region replication active with zero active VM instances in europe-west1, and configure automated failover scripts to submit regional Compute Engine quota increases and spin up MIG instances upon failure declaration.
  4. D
    Migrate the database layer from Cloud SQL to Cloud Spanner with a multi-region configuration spanning europe-west3 and europe-west1, and continuously run fully scaled active-active MIG clusters at 100% baseline compute capacity in both regions.

Cevap

Maintain a cross-region read replica of Cloud SQL in europe-west1, alongside a minimal Warm Standby MIG scaled to one instance using pre-validated instance templates. Upon DR declaration, promote the read replica to standalone primary, scale out the MIG to baseline capacity, and update Cloud DNS routing.
The correct strategy balances RPO, RTO, and cost by using Cloud SQL cross-region asynchronous read replication (yielding sub-minute RPO) alongside a Warm Standby MIG pre-configured with instance templates and minimal baseline instance count in europe-west1. Promoting the database replica and scaling the existing MIG allows operational recovery well under 15 minutes while keeping steady-state secondary compute costs extremely low.

Adım Adım Çözüm

1
Analyze RPO requirements
RPO < 1 minute requires continuous asynchronous data replication between regions rather than periodic batch file backups.
Cloud SQL cross-region read replicas continuously stream write-ahead logs (WAL) from europe-west3 to europe-west1, securing near-zero RPO.
2
Analyze RTO and cost constraints
RTO < 15 minutes paired with minimal idle cost favors a Warm Standby compute pattern over Cold Backup or fully scaled Hot Standby.
Maintaining a single minimal instance in the secondary MIG validates template readiness and quota availability, enabling rapid scale-out within minutes during failover.
3
Evaluate execution sequence
Promote Cloud SQL replica to primary, trigger MIG scale-out, and re-point DNS routing.
Promoting the database read replica creates a writeable database instance in europe-west1 while compute capacity expands to handle full production traffic.

Anahtar Kavram

Disaster Recovery Strategy Selection (Warm Standby vs. Cold Backup vs. Active-Active)
Bu soruyu puanla

Konular

Disaster Recovery and Business Continuity Execution

Bu konudan daha fazla soru