Host-Based Active/Active Data Replication Across Four Sites
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data replication systems in active/active storage environments are complex and prone to multiple failure points, particularly in four-site replication scenarios, which complicates data synchronization and error handling across geographically distant sites.
Innovation Solution
A method and system for data replication that uses a host to synchronously replicate data between two local storage devices and asynchronously replicate it to remote storage devices, allowing for operational control swapping between hosts to ensure consistent data replication across multiple sites, thereby simplifying the replication process and reducing complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If storage controllers at each remote destination maintain maps and control replication features, then data replication functionality is provided, but system complexity increases and multiple failure points are introduced
Solution Approach 1:
The patent extracts the intelligence and control functions from the storage controllers at remote destinations and relocates them to a central storage controller at the source site. This is achieved by having the source storage controller maintain replication maps and send control instructions to remote destinations, rather than distributing control intelligence across all nodes. This extraction of complexity from individual nodes to a central controller reduces system complexity while maintaining full replication functionality.
Solution Approach 2:
The patent merges the replication control functions from multiple distributed storage controllers into a single centralized storage controller. By combining the map-maintenance and control-feature capabilities at one location rather than distributing them across multiple sites, the system achieves simpler architecture with fewer failure points while preserving the ability to replicate data across four sites.
2Reliability
If synchronous transfers are used between geographically distant sites, then data consistency is maintained, but transfer time increases beyond preferable limits
Solution Approach 1:
The patent applies preliminary action by having the source storage controller pre-calculate and maintain replication maps that identify which data blocks need to be replicated to which remote destinations. This preparation of replication information in advance allows the system to efficiently initiate transfers only when necessary, avoiding continuous synchronous transfers across long distances while ensuring data consistency is maintained through targeted updates.
Data Source
AI summary
A system and method for data replication is provided. A host synchronously replicates data between a first local storage device and a second local storage device, the first local storage device and the second local storage device being coupled to the host. Data is then asynchronously replicated from the first local storage device and the second local storage device to at least one remote storage device. In an embodiment, each of the local storage devices may asynchronously replicate data to remote storage devices located at separate sites to provide four site data replication processing. The system described herein provides for advantageous and efficient swapping of control between hosts and storage devices located in different regions in response to disasters and/or other events affecting data storage in a particular region.