Data Replication Using Point-In-Time Copy for Continuous Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing replication systems face challenges in creating an initial copy of source data with minimal impact on application access, as they often require exclusive control of the source data, leading to extended downtime and potential disruptions.
Innovation Solution
Generating a point-in-time copy of the source data set and using a restored copy to create an initial replication, allowing continuous application access by avoiding serialization during the data transfer process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the replication system obtains exclusive control of the source data set during the refresh or initial copy stage, then the copied target data is point-in-time consistent with the source data, but application access to the source data is adversely affected
Solution Approach 1:
The system segments the data copying process by creating a point-in-time copy of the source data set and using a restored copy for the actual data transfer to the target. This segmentation allows the original source data to remain accessible to applications while a separate copy is being replicated, resolving the contradiction between maintaining data consistency and ensuring continuous application access.
Solution Approach 2:
The system creates a point-in-time copy of the source data set and uses this copy (restored copy) for the initial data transfer to the target. By working with the copy rather than the original source data, the system ensures that applications can continue to access the original source data without interruption while the replication process uses the restored copy, thus maintaining both data consistency and application availability.
2Reliability
If the refresh operation to create the initial copy is performed with exclusive control, then data consistency is ensured, but the operation takes an extended period of time
Solution Approach 1:
The replication process is segmented into creating a point-in-time copy and then using a restored copy for data transfer. This segmentation reduces the time the source data set needs to be locked, as the actual data copying operates on the restored copy rather than requiring continuous exclusive access to the original source, thereby reducing the extended downtime associated with traditional refresh operations.
Solution Approach 2:
The system performs a preliminary action by creating a point-in-time copy of the source data set before initiating the main data transfer process. This preliminary copy creation allows the subsequent data transfer to proceed using the restored copy, reducing the overall time impact on the source system while ensuring data consistency is captured at a specific point in time.
3Measurement precision
If exclusive control is obtained during the initial copy, then point-in-time consistency is achieved, but application access is blocked
Solution Approach 1:
The system creates a point-in-time copy of the source data set and uses this restored copy for the initial replication process. This approach achieves point-in-time consistency by capturing the source data state at a specific moment while allowing applications to continue accessing the original source data set without interruption, as the replication operates on the copy rather than the original.
Solution Approach 2:
The restored copy acts as an intermediary between the source data set and the target data store during the initial replication. This intermediary allows the system to achieve point-in-time consistency for replication purposes while the original source data remains accessible to applications, effectively mediating between the need for consistent data capture and continuous system availability.
Data Source
AI summary
Provided are a computer program product, system, and method for replicating a source data set to a target data store. A point-in-time copy of the source data set is generated having a data structure identifying the data in the source data set as of a point-in-time. A restore operation is initiated to copy the source data set represented by the point-in-time copy to a restored copy of the source data set consistent with the source data set. The source data set records are transferred from the restored copy to the target data store in the target storage.