Storage Array Controller Change Tracking for Data Replication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data replication techniques, such as file-level and volume-level replication, face drawbacks including the need for identical file systems at both sites, resource consumption by the production system, and unnecessary data transfer, which hinder efficient data reliability and availability.
Innovation Solution
Implementing a method for file-level replication using volume-level change tracking, where mapping information and change information are used to identify modified regions of a storage device, allowing replication to be initiated by a non-production system, thereby reducing resource overhead and enabling selective data replication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If file-level replication is used to track changes at the file level, then replication precision is improved, but the production system must consume resources to initiate replication, reducing productivity
Solution Approach 1:
The patent introduces a storage array controller as an intermediary that autonomously performs change tracking and replication initiation. The controller includes a change tracker that monitors volume changes and a replicator that initiates replication without production system involvement, resolving the contradiction by removing the burden from the production system while maintaining precise change tracking through the intermediary's dedicated functionality
Solution Approach 2:
The storage array controller performs self-service by autonomously tracking changes to volumes and initiating replication operations without requiring the production system to consume resources. The controller's change tracker and replicator work independently to monitor and replicate changes, allowing the production system to focus entirely on application services while the storage system handles replication self-servingly
2Ease of operation
If volume-level replication is used to simplify replication initiation, then ease of operation is improved, but unnecessary data is replicated, increasing loss of energy
Solution Approach 1:
The patent segments the replication process by introducing granular change tracking at the extent level within volumes. Instead of replicating entire volumes or relying on coarse file-level tracking, the change tracker identifies specific extents (data blocks) that have changed, allowing selective replication of only the necessary portions. This segmentation enables precise control over what data is replicated, reducing unnecessary network transfers while maintaining operational simplicity
Solution Approach 2:
The patent applies local quality by enabling different replication behaviors for different regions of storage. The change tracker monitors specific extents within volumes and identifies which local regions have changed. This allows the system to replicate only the specific changed extents rather than entire volumes or files, optimizing network bandwidth usage by transmitting only the necessary data portions with appropriate change tracking information
3Reliability
If volume-level replication is used to track all changes, then reliability is improved, but adaptability deteriorates because file system types and layouts cannot be different between sites
Solution Approach 1:
The patent changes the dimension of operation from the file system layer to the storage array layer. By implementing change tracking and replication control at the storage array level (using volume and extent information) rather than at the file system layer, the system decouples replication functionality from file system dependencies. This dimensional shift allows different file system types and layouts to coexist at different sites while maintaining reliable replication through the storage array's universal volume-based tracking mechanism
Data Source
AI summary
Data-object-level replication is performed based on storage-device-level change tracking. A method of replicating a data object (such as a file) involves accessing mapping information and change information. The mapping information identifies that one or more regions of a storage device (such as a logical volume) store portions of the data object. The change information identifies whether the regions of the storage device, which were identified by the mapping information, have been modified subsequent to a particular point in time. If the change information identifies that any of those regions of the storage device have been modified, replication of each modified one of the regions to a secondary system is initiated. Replication can be initiated from a system other than the production system.