Asynchronous Data Replication Coordination Time Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Asynchronous data replication environments, such as IBM's Global Mirror, face significant delays in consistency group coordination times due to the distance between primary storage devices, leading to unacceptable host system impacts, especially in active/active configurations where devices are located thousands of kilometers apart.
Innovation Solution
Implementing a method where each primary storage device calculates and waits for a specific time before executing a Start Increment command to synchronize the queuing of host I/O across devices, using consistency coordination modules to determine network latency and performance, ensuring all devices execute the command at substantially the same time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If primary storage devices are located at different sites to enable geographic redundancy and failover capability, then system reliability is improved, but coordination time increases due to distance
Solution Approach 1:
The patent applies preliminary action by having primary storage devices calculate and wait for a predetermined time period before executing the Start Increment command. This pre-calculation of execution time based on device distance and performance characteristics allows devices to synchronize their command execution despite being geographically distributed, thereby resolving the contradiction between geographic redundancy and coordination time.
2Adaptability or versatility
If primary storage devices are located far apart to provide geographic distribution, then adaptability is improved, but host I/O response time deteriorates due to increased coordination delays
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the execution time parameter based on the distance and performance characteristics of each primary storage device. Each device calculates its specific wait time before executing the Start Increment command, allowing the system to maintain synchronized operation across geographically distributed devices without impacting host I/O response time.
3Reliability
If consistency groups are formed frequently to maintain data consistency, then data consistency is improved, but coordination overhead increases proportionally
Solution Approach 1:
The patent reduces coordination overhead by applying preliminary action - each primary storage device pre-calculates its execution time for the Start Increment command based on its distance and performance characteristics. This allows devices to execute commands in a synchronized manner without requiring multiple coordination cycles, thereby maintaining data consistency while reducing the frequency and overhead of coordination operations.
Data Source
AI summary
A method for reducing coordination times in asynchronous data replication environments is disclosed. In one embodiment, such a method includes providing multiple primary storage devices in an asynchronous data replication environment. A command is issued, to each of the primary storage devices, to begin queuing I/O in order to coordinate a consistency group. Each primary storage device receives the command. The method further calculates, for each of the primary storage devices, an amount of time to wait before executing the command with the objective that each primary storage device executes the command at substantially the same time. Each primary storage device is configured to execute the command after receiving and waiting its corresponding amount of time. A corresponding system and computer program product are also disclosed herein.