Examkin

Soru

Zorluk: Çok zorDesigning Migration Plans and Data Transfer Strategies

A multinational retail enterprise is planning to migrate its core e-commerce database (a 40 TB PostgreSQL instance) and associated 300 TB object store from an on-premises data center to Google Cloud (Cloud SQL for PostgreSQL and Cloud Storage). The enterprise mandates continuous data synchronization prior to cutover to comply with a strict maximum allowed maintenance window of 30 minutes. What is the correct sequence of steps to plan and execute this zero-data-loss migration strategy?

  1. 1Establish a Cloud Interconnect with Private Google Access and configure Cloud Router BGP sessions to secure network connectivity.
  2. 2Perform a baseline data migration by running Storage Transfer Service for object storage and restoring a full database snapshot onto Cloud SQL for PostgreSQL.
  3. 3Configure Database Migration Service (DMS) with continuous replication (CDC) to stream ongoing transaction logs from the on-premises PostgreSQL source to Cloud SQL.
  4. 4Initiate the maintenance window, stop write traffic on the on-premises database, verify that replication lag reaches zero, and promote the Cloud SQL instance to primary.
  5. 5Reconfigure application connection strings to point to the new Cloud SQL primary endpoint, run health validation checks, and open traffic to end users.

Cevap

The correct execution sequence is: Establish hybrid network connectivity -> Perform baseline data and database snapshot transfer -> Configure continuous CDC via Database Migration Service -> Freeze on-premises writes, wait for zero replication lag, and promote target database -> Redirect application traffic to GCP endpoints and validate operational status.
A structured database migration designed for strict SLA maintenance windows must follow a logical dependency chain: 1) Secure network connectivity via Interconnect, 2) Bulk initial baseline transfer for large static storage and initial database snapshot, 3) Real-time CDC streaming via Database Migration Service to minimize divergence, 4) Controlled maintenance window freeze of source write operations, confirmation of zero replication lag, and promotion of target instance, and 5) Final application endpoint redirection and validation.

Adım Adım Çözüm

1
Establish network connectivity foundation
Hybrid connectivity via Dedicated or Partner Interconnect with Cloud Router BGP sessions established.
Secure, reliable, high-throughput private communication must precede any bulk or streaming data movement between environments.
2
Execute initial bulk baseline load
300 TB object data transferred via Storage Transfer Service and 40 TB database snapshot restored onto Cloud SQL target.
Establishes the initial historical data baseline in Google Cloud before starting delta capture.
3
Enable continuous Change Data Capture (CDC)
GCP Database Migration Service (DMS) actively streams logical write-ahead logs (WAL) from on-premises PostgreSQL to Cloud SQL.
Maintains low replication lag and near-real-time parity between source and target databases while production remains live.
4
Execute cutover and promote target instance
On-premises application writes are stopped, DMS sync achieves zero-second lag, continuous replication is stopped, and Cloud SQL target is promoted to standalone primary.
Guarantees zero data loss and prepares the target database to accept new write operations within the 30-minute maintenance window.
5
Redirect traffic and complete migration
Application configuration endpoints updated to target Cloud SQL and Cloud Storage; health checks pass.
Completes final transition of user requests to the GCP infrastructure.

Anahtar Kavram

Minimal-downtime database and data migration using baseline snapshot transfer and continuous CDC replication via GCP Database Migration Service.
Bu soruyu puanla

Konular

Designing Migration Plans and Data Transfer Strategies