Failover Backspacing Synchronization for Data Centers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for synchronizing data between distant computer centers in high-availability clusters, such as those used in financial transaction terminals, are unreliable and require significant hardware investment due to the time-critical nature of synchronization processes, especially in active/active operations where distances exceed 25 km.
Innovation Solution
A method where transaction terminals are temporarily taken out of service to synchronize configuration and transaction data between computer centers, with a signal indicating the primary center's operational status, followed by reactivation once synchronization is complete, allowing seamless switching back to the primary center without data loss.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional synchronization mechanisms are used between distant computer centers, then data synchronization can be achieved, but the process becomes unreliable and requires very great hardware expense
Solution Approach 1:
The patent applies preliminary action by detecting the operational status of the primary computer center before initiating failover or synchronization operations. The secondary center continuously monitors whether the primary center is operational, and only performs data synchronization or failover when necessary. This prevents unnecessary synchronization attempts that would fail due to network unreachability, thereby improving reliability without requiring expensive hardware modifications.
2Reliability
If data synchronization is performed between widely separated computer centers, then backup capability is maintained, but transmission time increases and transaction data may be lost
Solution Approach 1:
The patent applies dynamics by making the synchronization process adaptive to network conditions. The system dynamically adjusts its behavior based on whether the primary computer center is reachable: if unreachable, it prevents synchronization attempts that would cause data loss; if reachable, it performs normal synchronization. This dynamic approach maintains transaction data integrity while minimizing unnecessary delays caused by failed synchronization attempts.
Solution Approach 2:
The patent uses an intermediary mechanism in the form of a status detection step that mediates between the secondary center and the primary center. Before attempting synchronization, the secondary center detects whether the primary center is operational. This intermediary detection process prevents direct synchronization attempts that would fail due to network issues, thereby protecting transaction data integrity without incurring time losses from failed operations.
3Productivity
If the primary computer center is restored after failover, then service continuity is maintained, but unsafe backspacing can cause data loss
Solution Approach 1:
The patent applies feedback by implementing a controlled backspacing process that provides feedback before switching clients back to the primary center. The system performs a safety check to detect whether the primary center has been restored and is operational. Only after confirming the primary center's operational status does the system allow clients to switch back. This feedback mechanism ensures service continuity while preventing data loss that could occur from unsafe backspacing operations.
Data Source
AI summary
A method (100) is proposed for safely switching back to a first computer center (DC1) following failover by a second computer center (DC2), where at least one transaction terminal is connected as a client (CLA) to the second computer center (DC2) during the downtime of the first computer center (DC1) and at the end of the downtime is switched back again to the first computer center (DC1), where the following steps are performed: the second computer center (DC2) is shown that the first computer center (DC1) is operational again (step 110); the client (CLA) is prompted to go out of service (step 120); synchronization (STD) of configuration and/or transaction data between the first and second computer centers is performed (step 130); at the conclusion of synchronization(STD), the client (CLA) is prompted to switch back to the first computer center (DC1) (step 140).