Geo-Redundancy
Definition
Geo-Redundancy involves storing copies of data in multiple geographic locations, ensuring data availability and protection against regional disasters.
Use Cases
- Netflix: Maintain availability of streaming services during a regional cloud outage by keeping critical data and services resilient across regions. — Netflix is known for multi-region resilience practices on AWS, using regional isolation and disaster recovery patterns; for data, organizations commonly pair these patterns with cross-region replication for object storage and backups so content metadata and recovery artifacts remain available if a region is impaired. (Improved resilience to regional failures and reduced risk of prolonged service disruption; supports business continuity goals during large-scale incidents.)
- Dropbox: Protect user files from data center or regional disasters while maintaining durability expectations for stored content. — Dropbox has publicly described using geographically distributed infrastructure and replication to increase durability and availability of stored files, ensuring copies exist in separate locations to mitigate localized disasters. (Higher data durability and improved disaster recovery posture, reducing the likelihood of permanent data loss from a single-site event.)
Provider Equivalents
- AWS: Amazon S3 Cross-Region Replication (CRR)
- Azure: Azure Storage geo-redundant storage (GRS) / read-access geo-redundant storage (RA-GRS)
- GCP: Google Cloud Storage dual-region buckets
- OCI: OCI Object Storage Replication
Frequently Asked Questions
- What's the difference between Geo-Redundancy and multi-zone redundancy?
- Multi-zone redundancy keeps copies in different availability zones within the same region, protecting against a single data center failure. Geo-redundancy keeps copies in different regions (separate geographies), protecting against regional disasters like major power grid failures, earthquakes, or widespread network outages.
- When should I use Geo-Redundancy?
- Use geo-redundancy when you need disaster recovery from a full regional outage, have strict business continuity requirements (RTO/RPO targets), or must meet durability expectations for critical data (backups, customer records, essential application artifacts). If your main concern is a single data center failure and you want lower cost/latency, multi-zone redundancy in one region may be enough.
- How much does Geo-Redundancy cost?
- Costs typically include (1) extra storage for the replicated copy, (2) replication or data transfer charges between regions (varies by provider and direction), and (3) additional request/operation costs in the destination. Read-access options (like Azure RA-GRS) can add cost. The biggest drivers are data volume, change rate (how much new/updated data replicates), and how often you read from the secondary region during tests or failovers.
Category: data
Difficulty: intermediate
Related Terms
See Also