Intrigued by the art of cloud architecture? Discover how to design, develop, and manage robust, secure, scalable, and dynamic solutions on Google Cloud as you prepare for the Professional Cloud Architect exam!
Disaster Recovery (DR) is a set of processes designed to restore IT systems after a major failure or outage. Designing a robust architecture requires balancing two critical metrics with cost and complexity to ensure business continuity. The Recovery Time Objective (RTO) measures the maximum acceptable downtime, while the Recovery Point Objective (RPO) measures the maximum tolerable data loss.
To protect raw data against regional outages, Cloud Storage provides an active-active architecture across multiple locations using dual-regions and multi-regions. Administrators can enable Turbo Replication to sync data between regions within 15 minutes, which helps achieve a lower RPO. For bucket-level security, Soft Delete recovers accidentally deleted objects, Bucket Lock enforces immutable retention policies, and Object Holds place metadata flags on specific files to prevent them from being overwritten.
For managed database services, data protection involves a mix of automated backups and Point-in-Time Recovery (PITR). Databases like Firestore and Spanner support both scheduled backups for long-term retention of up to one year and PITR to restore data to any specific minute within the last seven days. To minimize downtime, administrators can deploy regional databases across three zones for 99.99% availability, or multi-region databases across five zones in three regions for 99.999% availability.
Analytical systems like BigQuery require specialized strategies to handle massive datasets during a disaster. While BigQuery provides built-in redundancy, exporting tables to Cloud Storage in a different region is a best practice for long-term durability. To recover from user errors, Time Travel lets users query data as it existed up to seven days ago, while the Fail-safe Period provides an extra seven days of recovery through Google support. To bring backup data from on-premises environments into the cloud safely, the Storage Transfer Service automates scalable data moves without the risks of manual scripting.
An effective disaster recovery plan requires periodic validation of recovery workflows to ensure the system behaves as expected during an actual crisis. Administrators can combine the gcloud CLI with Cloud Scheduler to automate routine restoration tests and keep configurations consistent. For critical workloads, cross-region replication should be configured so that if the primary region fails, traffic and operations failover to a secondary location with minimal impact.
Cloud Storage provides highly durable object storage designed with 11 9's of durability to protect files from physical hardware failures. While the underlying infrastructure is resilient, human error remains a common cause of data loss. To prevent accidental deletion, administrators can configure Soft Delete to retain deleted items for a recovery window, or enable Object Versioning to maintain a history of object states whenever files are overwritten.
Meeting regulatory and compliance standards requires strict data retention controls that cannot be bypassed. Google Cloud provides Bucket Lock to set a minimum retention period for all objects in a bucket, and Object Retention Lock to manage retention requirements on a per-object basis for maximum control. To manage costs and satisfy legal requirements over time, Lifecycle Management Policies can be configured to automatically transition older data to cheaper archive tiers or permanently delete it once the retention period ends.
To protect business operations against massive geographic disasters, data can be stored across different regions. Setting up dual-regions or multi-regions ensures that data is stored in multiple physical locations. For dual-region configurations, Turbo replication speeds up syncing by replicating objects between regions within a target 15-minute window.
Moving data into Cloud Storage for backup purposes is managed through specialized ingestion tools. The Storage Transfer Service handles high-volume data transfers between cloud providers or from on-premises systems, while the Backup and DR Service offers a centralized console to manage backups for virtual machines and databases. For massive physical migrations where network bandwidth is a constraint, a Transfer Appliance can be shipped directly to a data center to move data to the cloud offline.
For virtual machine instances, Persistent Disks provide block storage that requires consistent protection to maintain business continuity. To capture the state of these disks, administrators create snapshots, which are incremental backups that only record data changed since the last backup to reduce storage costs. These scheduled snapshots are configured to run automatically at set intervals, and they can be managed centrally alongside virtual machine backups using the Backup and DR Service.
Relational database systems require low RPO and RTO strategies to prevent corruption and handle immediate failovers. Both Cloud SQL and AlloyDB use Point-in-Time Recovery (PITR), allowing administrators to restore database states to a specific second to achieve a near-zero RPO. AlloyDB also supports continuous backups that run in the background without affecting database performance. For high availability, these databases use standby instances in different zones for instant failover, alongside cross-region replicas to protect against regional outages.
NoSQL services such as Spanner and Firestore use scheduled backups and recovery features to secure document and transactional databases. Spanner supports multi-regional configurations that automatically replicate data across different geographic areas to protect against a total regional outage. Firestore enables users to perform Surgical Recovery of specific historical data by reading documents from up to seven days in the past. To prevent malicious or accidental administrative actions, users can enable Database Deletion Protection to block the removal of entire databases.
For long-term retention of database exports, Cloud Storage serves as a secure and highly durable repository. Inside these backup buckets, administrators can lock down files using Object Holds and Bucket Lock to enforce immutable retention periods. Furthermore, Cross-Bucket Replication can be configured to automatically duplicate backup archives to a secondary, independent storage bucket for an extra layer of redundancy.
Prepare and test your skills
Prepare and test your skills
BigQuery Time Travel allows users to directly query data as it existed up to seven days ago to recover from user errors. In contrast, the Fail-safe Period provides an extra seven days of data recovery that can only be accessed through Google support.
Bucket Lock enforces a minimum retention period across all objects in an entire bucket to maintain immutable retention policies. In contrast, Object Retention Lock manages compliance retention requirements on a per-object basis for more granular control.
Turbo Replication reduces the Recovery Point Objective (RPO) by syncing data between dual-region locations within a target 15-minute window. This accelerated replication ensures minimal data loss if an outage affects the primary storage location.
Firestore Point-in-Time Recovery (PITR) restores data to any specific minute within the last seven days, whereas Cloud SQL and AlloyDB allow restoration to any specific second. Restoring to a specific second enables relational databases to achieve a near-zero Recovery Point Objective (RPO).
Enable Object Versioning on the bucket, configure an Object Lifecycle Management rule to delete noncurrent versions after seven years, and enforce public access prevention.
Configure a bucket retention policy of seven years, lock the retention policy using Bucket Lock, and configure an Object Lifecycle Management rule to delete objects older than seven years.
Place an event-based hold on all bucket objects and configure a Cloud Key Management Service (Cloud KMS) customer-managed encryption key with scheduled key destruction after seven years.
Configure Object Retention Lock in Unlocked mode for each object and set a Soft Delete retention duration of seven years on the bucket.
A financial services organization is migrating transaction archive data to Google Cloud Storage to satisfy SEC and FINRA regulatory compliance mandates. The compliance team outlines the following mandatory storage requirements:
Which Cloud Storage configuration should you implement to satisfy these requirements?
StudyCerts
StudyCerts