Autonomous Service Node Switching in Distributed Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current distributed storage systems have low metadata service recovery efficiency due to inadequate service node switching mechanisms, which fail to address abnormal states such as slow responses or network issues, leading to poor user experience.
Innovation Solution
A method and apparatus for monitoring service node response states through multithreading, stopping communication with abnormal nodes, and triggering a switching process to select a new service node, using abnormal node flags to prevent misdiagnosis and ensure stable recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If centralized metadata management with heartbeat mechanism is used, then system architecture is simple and easy to operate, but service node switching is delayed and metadata service recovery efficiency is low
Solution Approach 1:
The patent implements a feedback mechanism where backup service nodes actively monitor the current service node's responsiveness to service requests. When the current service node fails to respond within a threshold time, the backup nodes detect this abnormality and trigger switching. This feedback loop enables timely detection and response to service node failures, resolving the contradiction between simple operation and efficient recovery.
Solution Approach 2:
The system enables self-service through automated service node switching. When abnormalities are detected, the backup service nodes automatically perform switching operations without manual intervention. This self-service mechanism improves recovery efficiency while maintaining operational simplicity, as the system autonomously handles failure recovery.
2Device complexity
If heartbeat acknowledgment mechanism is used for node monitoring, then system complexity is reduced, but abnormal states like slow responses or network issues are not detected, leading to failed service node switching
Solution Approach 1:
The patent applies preliminary action by having backup service nodes proactively send service requests to monitor the current service node's state before actual failures occur. This preliminary monitoring detects slow responses and network issues early, enabling reliable service node switching while maintaining relatively simple system complexity through the use of existing service request mechanisms.
3Productivity
If service node switching is triggered only on complete unresponsiveness, then false switching is reduced, but service quality deteriorates due to delayed switching in partial failure states
Solution Approach 1:
The patent changes the monitoring parameter from binary heartbeat acknowledgment to timed service request response. By measuring the response time of service requests and comparing it against a threshold, the system can detect partial failures like slow responses. This parameter change enables faster switching in degraded states while maintaining service quality, resolving the contradiction between switching speed and service quality.
Data Source
AI summary
Provided in the disclosed embodiments are methods and apparatuses for switching service nodes in a distributed storage system, the service nodes comprising a current service node and backup service nodes. The method comprises: monitoring response states of the service nodes to a service request; and if a response state of the current service node is abnormal, stopping communication between the current service node and the backup service nodes, and triggering a switching process for the current service node. In the case of a service timeout, service unavailability, service abnormality, or the like caused by a hardware failure or software defect, logic is checked with a service node that performs logical determination and data statistics collection for multiple factors affecting response states of service nodes, thereby achieving autonomous switching and recovery of the service nodes and improving service availability.