Remote Copy State Management for Network Disconnection Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems for remote copy face challenges in maintaining data redundancy during network or device failures in unstable power and network environments, where immediate purging is not feasible due to unreliable infrastructure.
Innovation Solution
Implement a method that temporarily suspends the state of copy pairs during network disconnection, detects network recovery, and resumes operations only when recovery is confirmed within a certain period, or proceeds with failover if recovery is not detected within the threshold time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If immediate purging is performed upon network disconnection detection, then system reliability is improved by quickly transitioning to failover mode, but data redundancy is lost and application availability deteriorates in unstable network environments where temporary disconnections occur
Solution Approach 1:
The system performs preliminary actions by transitioning to a temporary suspended state before final failover, allowing it to detect and respond to network recovery. This intermediate state enables the system to prepare for potential recovery without immediately committing to failover, thus preserving data redundancy opportunities while maintaining reliability.
Solution Approach 2:
The system dynamically adjusts its response based on network conditions by introducing a threshold period for recovery detection. Instead of a static immediate failover response, the system can transition between temporary suspended and suspended states based on whether network recovery occurs within the threshold period, optimizing both reliability and application availability.
2Productivity
If the system waits for network recovery confirmation before failover, then data redundancy is maintained and application availability is improved, but system reliability deteriorates due to prolonged suspension in case of permanent network failures
Solution Approach 1:
The system performs preliminary network recovery detection within a threshold period before committing to failover. This preliminary check allows the system to maintain data redundancy when appropriate while having a clear timeout mechanism to prevent prolonged suspension, thus balancing application availability with system reliability.
Solution Approach 2:
The system implements feedback by continuously monitoring network connection status during the temporary suspended state. If network recovery is detected within the threshold period, the system transitions back to copy pair state; otherwise, it proceeds to suspended state. This feedback loop ensures optimal decisions based on actual network conditions, maintaining both availability and reliability.
3Reliability
If the threshold period for network recovery detection is extended, then data redundancy is better preserved by allowing more time for recovery, but loss of time increases due to prolonged suspension before failover execution
Solution Approach 1:
The system uses parameter changes by adjusting the threshold period duration to balance data redundancy preservation with downtime minimization. The threshold period is configured as a可调 parameter that can be optimized based on specific network stability requirements, allowing the system to adapt to different operational contexts while managing the trade-off between redundancy and time loss.
4Device complexity
If the system immediately transitions to suspended state upon network disconnection, then system complexity is reduced by simplifying the state transition logic, but adaptability deteriorates as the system cannot respond to temporary network interruptions
Solution Approach 1:
The system segments the state transition logic into distinct states: copy pair state, temporary suspended state, and suspended state. This segmentation allows the system to handle different network conditions appropriately while maintaining manageable complexity. The temporary suspended state acts as an intermediate buffer that provides adaptability without significantly increasing overall system complexity.
Data Source
AI summary
While a connection between a copy pair comprising a primary volume in a first storage system associated with a primary host and a secondary volume in a second storage system associated with a secondary host is disconnected, systems and methods described herein can involve changing a state of the copy pair for each of the primary volume and the secondary volume to a temporary suspended state; detecting whether a network connection has recovered; for the detecting indicative of the network connection having recovered, changing the state of the copy pair for the each of the primary volume and the secondary volume to a copy pair state; and for the detecting indicative of the network connection not being recovered after a threshold period of time has elapsed, changing the state of the copy pair for the each of the primary volume and the secondary volume to a suspended state.