Active-Active Data Center Self-Arbitration After Arbitration Device Failure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In active-active data centers, when the arbitration device fails, there is no mechanism to determine which data center should continue providing service, leading to simultaneous service stoppage and data inconsistency.
Innovation Solution
Data centers implement a preset arbitration policy to determine if they are preferred, allowing the preferred data center to continue service when communication is interrupted and the arbitration device is faulty.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If an arbitration device is used to determine service continuation, then data consistency is ensured, but service interruption occurs when the arbitration device fails
Solution Approach 1:
The data centers perform self-arbitration by autonomously determining service continuation status based on preset arbitration policies when communication with the arbitration device is interrupted, eliminating dependency on the arbitration device for service continuity
Solution Approach 2:
Arbitration policies are configured in advance in the data centers, enabling them to make immediate arbitration decisions without external intervention when the arbitration device fails, thus preventing service interruption
2Productivity
If both data centers continue service when communication is interrupted, then service continuity is maintained, but data inconsistency occurs
Solution Approach 1:
Each data center independently evaluates its own service continuation status by applying arbitration policies locally, ensuring that only one data center continues service and data consistency is maintained without requiring real-time coordination with the other data center
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
Embodiments of the present invention provide an arbitration method, apparatus, and system that are used in active-active data centers, and relate to the field of computer technologies. The arbitration method includes: when a first data center determines that communication between the first data center and a second data center is interrupted, and an arbitration device cannot perform arbitration, viewing, by the first data center, a preset arbitration policy; and continuing providing a service when determining, according to the arbitration policy, that the first data center is a preferred data center; or stopping providing a service when determining, according to the arbitration policy, that the first data center is not a preferred data center; where the first data center and the second data center are the active-active data centers. According to the foregoing solution in the present invention, when the first data center determines that the communication between the first data center and the second data center is interrupted, and the arbitration device cannot perform arbitration, only the preferred data center continues providing the service, which solves a problem of a service interruption caused by a fault of the arbitration device, and ensures that an uninterrupted service is provided for a user.