Virtualizing Remote Copy Pairs in Three Data Center Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In active-active high availability (HA) storage systems, existing technologies fail to correctly recognize and manage remote copy pairs across three data centers, leading to suspension of remote copy data protection when a storage system failure occurs, as the server cannot differentiate between active and standby pairs, resulting in continued IO processing and inability to change remote copy pair status.
Innovation Solution
The solution involves virtualizing remote copy pairs by using virtual volume identifiers to manage copy pairs across three data centers, allowing clustering software to automate the activation of standby pairs upon failure, ensuring data consistency and resynchronization, and enabling the server to correctly manage and switch between active and standby pairs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional multi-target remote copy system is used, then the server can access only a first volume, but when the first storage system failure occurs, the server cannot recognize the failure and continue processing IO, causing suspension of remote copy data protection
Solution Approach 1:
The patent segments the remote copy pair identification into two distinct layers: virtual volume identifiers presented to the server, and physical copy pair identifiers used internally by storage systems. This segmentation allows the server to interact with a simplified view while the underlying system maintains detailed tracking of multiple physical copy pairs, enabling proper failure detection and status management without complicating server operations.
Solution Approach 2:
The patent introduces an intermediary mapping mechanism between virtual volume identifiers and physical copy pair identifiers. This intermediary layer translates server requests into appropriate physical copy pair operations, allowing the server to issue commands without needing to understand or differentiate between multiple physical copy pairs. The intermediary maintains the mapping relationships and routes operations to the correct physical copy pairs, enabling automated failover and status changes.
2Productivity
If physical system configuration with two remote copy pairs is used, then data can be copied to third storage system, but the server recognizes them as one volume and cannot control and operate the remote copy system correctly
Solution Approach 1:
The patent merges the management of multiple physical copy pairs under a single virtual volume identifier from the server's perspective. The server interacts with one unified virtual volume while the system internally manages multiple physical copy pairs (first copy pair from first storage system, second copy pair from second storage system). This merging simplifies server operations while maintaining the productivity benefits of multiple physical copy pairs for data copying and redundancy.
Solution Approach 2:
The virtual volume identifier serves multiple functions: it presents a unified interface to the server, masks the underlying complexity of multiple physical copy pairs, and enables automated failover between different storage systems. The same virtual volume identifier can map to different physical copy pairs depending on failover status, providing multi-functionality that simplifies both data copying and system management.
3Adaptability or versatility
If active-active configuration is used, then IO can be processed in second storage system after first storage system failure, but the server cannot recognize the failure and there is no chance to change the remote copy pair status
Solution Approach 1:
The patent implements feedback mechanisms where the storage systems monitor each other's status and automatically detect failures. When the first storage system fails, the second storage system receives feedback about the failure condition and automatically initiates failover procedures. The system continuously monitors copy pair statuses and provides feedback to update the mapping between virtual volume identifiers and physical copy pair identifiers, enabling automated status changes without server intervention.
Solution Approach 2:
The patent establishes preliminary mappings between virtual volume identifiers and physical copy pair identifiers before failures occur. These pre-configured mappings include backup relationships between first and second storage systems. When a failure is detected, the system executes pre-planned failover actions by switching the virtual volume identifier mapping to point to the standby physical copy pair, enabling rapid automated response without requiring the server to recognize or initiate the failure response.
Data Source
AI summary
A first storage system manages a first volume, with a first internal volume identifier, and to provide the first volume with a given first virtual volume identifier to a server. A second storage system manages a second volume, with a second internal volume identifier, and to provide the second volume with the given first virtual volume identifier to the server. A third storage system manages a third volume. The first storage system manages a first copy pair between the first volume and the third volume with a first internal copy identifier, and to provide a first copy pair with a given virtual copy pair identifier to the server. The second storage system manages a second copy pair between the second volume and the third volume with a second internal copy identifier, and to provide a second copy pair with the given virtual copy pair identifier to the server.