Network Edge Exception Detection for Millisecond Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current redundancy systems face challenges in accurately and quickly detecting network exceptions between data centers, leading to prolonged detection times and service disruptions due to unstable network connections, especially when external factors cause network instability.
Innovation Solution
Implementing a device management method where a network edge device directly reports network exception packets to a first device, allowing for millisecond-level network switching to a backup network, using hardware-level monitoring and sequence number management to ensure data consistency and reduce service impact.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If statistical threshold-based detection is used to avoid misjudgment, then false positives are reduced, but detection accuracy decreases and detection time increases
Solution Approach 1:
The patent segments the detection function by introducing a dedicated network edge device that monitors network status independently from the storage devices. This edge device collects network indicators and generates exception packets when thresholds are exceeded, separating the detection logic from the data processing logic and enabling faster, more accurate detection without impacting storage operations.
Solution Approach 2:
The network edge device acts as an intermediary between the network and storage devices. It monitors network status, detects exceptions based on statistical thresholds, and communicates with storage devices through exception packets. This intermediary role allows the system to maintain high detection accuracy while minimizing the time storage devices spend on detection activities.
2Stability of the object's composition
If large statistical threshold is used to avoid unnecessary network switching, then false switching is reduced, but detection sensitivity decreases
Solution Approach 1:
The patent applies different threshold levels for different network indicators (e.g., packet loss rate, delay, jitter) based on their specific characteristics and impact on service quality. Each indicator has its own optimized threshold that balances false positive reduction with detection sensitivity, allowing the system to maintain stability while improving detection precision for each specific metric.
3Adaptability or versatility
If network exception detection waits for timeout, then protocol compatibility is maintained, but service disruption duration increases
Solution Approach 1:
The network edge device continuously monitors network status and performs preliminary detection before actual service disruption occurs. By detecting exceptions based on statistical thresholds and proactively notifying storage devices through exception packets, the system triggers network switching before timeout conditions develop, maintaining protocol compatibility while significantly reducing service disruption duration.
Data Source
AI summary
A device management device and method, the method including receiving, by the first device, a network exception packet from a network edge device, where the network exception packet indicates that a network status of the first network is abnormal, the first device and a second device perform transmission of a service packet via the first network, and the first network comprises the network edge device connected to the first device, and switching, by the first device, in response to receiving the network exception packet, to transmitting a backup packet to the second device via a second network, where the second network is a transmission network whose network status is normal between the first device and the second device, and the backup packet comprises a backup packet corresponding to a service packet transmitted in the first network when the first network is abnormal.