Distributed Point-in-Time Copy for Consistent Disaster Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data synchronization methods in distributed systems face challenges in maintaining data consistency and integrity across multiple sites, particularly in disaster recovery scenarios, due to asynchronous data transfer and inconsistent snapshot processes, which can lead to application errors and performance degradation.
Innovation Solution
A method is provided which involves a controller component verifying the state of all file sets, and a sub-controller component suspends write operations to all databases at the primary site computer. Based on the suspending completing without error, the sub-controller executes peer snapshot processes for each of the databases and for application non-database files at the primary site computer. The applications using the databases and non-database files are not suspended in order to execute the peer snapshots. The peer snapshot copies can be executed in parallel or serially, and they effectively suspend writes by the application until they finish so that all data is consistent at the time of the copy. After the peer snapshot copy processes finish without error, each of the peer snapshots are automatically queued for transmission to a secondary site.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If asynchronous data transfer is used for disaster recovery, then data availability at secondary site is improved, but data consistency and integrity deteriorate
Solution Approach 1:
The system performs a preliminary suspension of write operations to all databases at the primary site before executing peer snapshot processes. This preliminary action ensures that no new data changes occur during the snapshot process, guaranteeing data consistency and integrity in the copied data set that is transferred to the secondary site.
Solution Approach 2:
The system creates peer snapshots (copies) of databases and application non-database files at the primary site. These peer snapshots are then transmitted to the secondary site, providing a consistent copy of the data that can be used for disaster recovery without requiring continuous synchronization.
2Loss of information
If write operations are suspended during snapshot processes, then data consistency is improved, but application performance deteriorates
Solution Approach 1:
The system extracts and suspends write operations only to databases at the primary site, while leaving application write operations to non-database files unaffected. This selective approach ensures data consistency for copied files while minimizing impact on application performance by allowing applications to continue writing to non-database files.
Solution Approach 2:
The system applies the write suspension only partially - specifically to databases that need to be copied - rather than suspending all write operations system-wide. This partial action achieves the necessary data consistency for the snapshot while maintaining application performance for non-database operations.
3Loss of information
If all applications are suspended during data copy, then data consistency is improved, but system productivity deteriorates
Solution Approach 1:
The system segments the data copy process into separate peer snapshot processes for each database and application non-database files. Applications are not required to be suspended globally; instead, only specific write operations to databases are suspended temporarily, allowing other applications to continue running and maintaining overall system productivity.
Solution Approach 2:
The system performs preliminary verification of the state of all file sets before initiating the snapshot process. This preliminary action ensures that the copy can proceed with minimal disruption, and the brief suspension of database writes is sufficient to ensure consistency without requiring prolonged application suspension.
4Productivity
If peer snapshots are executed in parallel, then productivity is improved, but complexity of control increases
Solution Approach 1:
The system introduces a controller component as an intermediary that coordinates and manages multiple peer snapshot processes. This controller verifies the state of file sets, initiates sub-controller components, and manages the parallel execution of snapshots, providing centralized control that simplifies the management of parallel processes while maintaining high productivity.
Solution Approach 2:
The system implements feedback mechanisms where the controller monitors the execution status of peer snapshot processes and adjusts coordination accordingly. This feedback-based control manages the complexity of parallel execution by dynamically responding to the state of ongoing snapshot operations.
Data Source
AI summary
Method, computer program product, and computer system are provided. A controller component verifies the state of all file sets. The controller component initiates a sub-controller component, based on successfully verifying the state of the file sets. The sub-controller component suspends write operations to all databases at the primary site computer. Based on the suspending completing without error, the sub-controller executes point-in-time copy processes for each of the databases and for application non-database files at the primary site computer. Applications using the databases and non-database files are not suspended. Based on all of the point-in-time copy processes completing without error, the sub-controller component resumes write operations to all databases at the primary site computer. The point-in-time copies are automatically sent to the secondary site computer so that the applications can operate at the secondary site computer if the primary site computer becomes unavailable.