How to Keep Data Safe During a Site Failover

A method for keeping data backups synchronized and accessible even when a primary data center goes offline and systems must switch to a backup location.

What it covers

This patent describes a way to manage data recovery when a primary production site fails. It uses a star-like topology where data is replicated to a secondary site and backed up to a separate storage location. When a failover occurs, the system reverses the flow of data, treating the replication site as the new production site. It uses a delta marker stream to track differences between storage locations, ensuring that the backup site remains consistent with the new production state even during complex recovery operations.

What it doesn't cover

• —Does not cover simple data mirroring without a failover mechanism.
• —Does not cover data backup methods that lack a delta marker stream for tracking changes.
• —Does not cover cloud-native object storage replication that does not involve a site-based failover.
• —Does not cover hardware-agnostic data recovery that ignores the state of the production site storage.

The clever bit

The system uses a unified delta marker stream to manage differences across multiple storage tiers during a failover, effectively 'reversing' the replication logic to maintain data integrity without manual intervention.

Why it matters

In enterprise IT, maintaining business continuity is critical. This patent provides a specific orchestration logic for EMC's (now Dell Technologies) data protection ecosystem, ensuring that when a disaster occurs, the backup chain does not break. It is a foundational piece of logic for high-availability storage arrays used by large corporations to prevent data loss during site outages.

Real-world examples

• 1.Dell EMC RecoverPoint
• 2.Enterprise disaster recovery storage clusters
• 3.Multi-site data center synchronization systems