Data Recovery via Batch and Transaction ID Cataloging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data replication becomes increasingly difficult as database sizes grow, and service interruptions at central processing stations, such as power outages or natural disasters, lead to lag times in mirroring data, making it challenging to determine which data was successfully mirrored before the interruption, necessitating the reconstruction of all data to ensure no loss.
Innovation Solution
An apparatus comprising a capture site, central processing station, and disaster recovery station, where the capture site stamps and transfers paper instructions to create electronic data records with batch and transaction identification numbers, which are stored and transmitted to the central processing station, then to the disaster recovery station, allowing for identification and re-transmission of missed data records after a service interruption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is mirrored from central data processing station to safe site, then data preservation during service interruption is improved, but lag time in mirroring makes it difficult to determine which data was successfully mirrored
Solution Approach 1:
The patent applies preliminary action by pre-cataloging data records with unique identifiers (hash values, sequence numbers, timestamps) at the capture site before transmission to the central processing station. This preliminary cataloging enables the system to quickly determine which records were successfully processed and which need recovery, eliminating the lag time problem in identifying mirrored data during service interruptions.
Solution Approach 2:
The patent implements feedback mechanisms where the central processing station sends acknowledgments back to the capture site confirming successful data processing. The system maintains a catalog of processed records with unique identifiers, allowing the capture site to identify which records need retransmission after service interruption, thus resolving the uncertainty about which data was successfully mirrored.
2Reliability
If all data is reconstructed to ensure no data loss, then data integrity is improved, but time and resources required for reconstruction increase
Solution Approach 1:
The patent extracts only the essential information needed for data recovery - unique identifiers (hash values, sequence numbers, batch numbers) and metadata - and stores them in a catalog at the capture site. This extraction principle allows the system to quickly identify which specific records need recovery without having to reconstruct or verify all data, significantly reducing recovery time while maintaining data integrity.
Solution Approach 2:
The system performs preliminary cataloging of all data records with unique identifiers before service interruption occurs. This advance preparation enables rapid identification of missing records after interruption, eliminating the need for time-consuming full data reconstruction and allowing selective recovery of only the necessary records.
3Reliability
If data replication is performed for growing databases, then data safety is improved, but complexity of replication process increases
Solution Approach 1:
The patent segments the data replication process into distinct functional components: capture site for cataloging, central processing station for execution, and disaster recovery site for backup. Each component handles specific tasks independently, managing complexity through modular design while maintaining data safety across the distributed system.
Solution Approach 2:
The patent introduces an intermediary catalog structure that mediates between the capture site and central processing station. This catalog stores unique identifiers and metadata, acting as an intermediary layer that simplifies the replication process by providing a standardized interface for tracking and recovering data without requiring complex point-to-point communication between all system components.
Data Source
AI summary
A method for data processing comprising may include receiving a paper instruction at a CS and stamping the paper instruction with a predetermined batch number at the CS. The method may further include transferring the paper instruction to a data element at the CS and constructing an executable electronic data record. The method may include creating a transaction identification number for the record and appending the transaction identification number to the record. The method may include transmitting the record from the RS to a CPS and receiving the record at the CPS. The method may include storing the transaction identification number and the batch identification number of the record in a CPS-table of records and executing the record at the CPS. The method may include transmitting the executed record from the CPS to a DRS, following a discrete lapse of time from the receipt of the record.