SDNAS Synchronous Replication for Zero Data Loss Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current disaster recovery solutions for Software Defined Network Attached Storage (SDNAS) lack effective mechanisms for ensuring data availability and minimizing downtime during failures, particularly in scenarios where synchronous data replication is necessary for maintaining data integrity and accessibility.
Innovation Solution
Implementing synchronous data replication between primary and backup storage arrays, where user filesystems and Virtual Data Mover configurations are replicated, allowing for seamless failover and ensuring that both sites have identical metadata and identifiers, enabling zero data loss and uninterrupted access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If synchronous data replication is implemented between primary and backup storage arrays, then data integrity and availability are improved, but system complexity and implementation difficulty increase
Solution Approach 1:
The patent establishes synchronous replication relationships in advance between primary and backup storage arrays, creating remote data facilities before failures occur. This preliminary setup ensures that backup arrays are already synchronized and ready to immediately take over upon failure, eliminating the need for complex post-failure data synchronization and reducing implementation complexity.
Solution Approach 2:
The patent creates exact copies of user filesystems and configuration filesystems from the primary storage array to the backup storage array through synchronous replication. These copies include all metadata, identifiers, and data structures, ensuring complete data integrity while using proven replication mechanisms rather than inventing new complex systems.
2Reliability
If complete filesystem replication including metadata is performed, then failover reliability is improved, but replication time and processing overhead increase
Solution Approach 1:
The patent implements continuous synchronous replication where metadata and filesystem data are replicated in real-time as changes occur on the primary storage array. This continuous process ensures that the backup array maintains an up-to-date copy without requiring periodic batch operations, eliminating replication delays and ensuring immediate failover capability.
Solution Approach 2:
The patent merges the replication of user filesystems with the replication of configuration filesystems containing metadata, identifiers, and VDM configurations into a unified synchronous replication process. This consolidation ensures that all necessary data and control information are replicated together atomically, maintaining consistency without requiring separate replication passes that would increase time overhead.
3Productivity
If selective failover of individual filesystems or VDMs is enabled, then service continuity is improved, but control and management complexity increase
Solution Approach 1:
The patent segments the storage system into independent failover units including individual filesystems, Virtual Data Movers (VDMs), and storage arrays. This segmentation allows selective failover of only the affected components while keeping other services running on the primary array, maintaining service continuity without requiring complete system failover. The modular structure simplifies management by allowing granular control over failover scope.
Data Source
AI summary
Disaster recovery is provided for Software Defined Network Attached Storage where the data backup is implemented synchronous data replication. A synchronous replication remote data facility is created between a primary storage array and a backup storage array. User filesystems being handled by a virtual data mover (VDM) and a configuration filesystem describing the location of the user filesystems on the primary storage array are replicated on the remote data facility to the backup storage array. Individual filesystems, individual VDMs, or groups of VDMs may be selectively failed over from the primary storage array to the backup storage array.