A fault recovery method and electronic device

By segmenting and managing hot and cold data in the storage system and optimizing the silent mechanism, the problem of long-term silence during failover was solved, enabling rapid fault recovery and business continuity.

CN122111742BActive Publication Date: 2026-07-10INSPUR SUZHOU INTELLIGENT TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
INSPUR SUZHOU INTELLIGENT TECH CO LTD
Filing Date
2026-04-30
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

During failover, existing storage systems suffer from slow metadata flushing due to the global nature and coarse-grained strategy of the silent mechanism. This results in prolonged service interruptions for the entire storage cluster, impacting business continuity and user experience.

Method used

Each storage volume in the storage system is segmented by logical block address to determine the hot and cold priority of the segment. When the controller node fails, a hot and cold data eviction strategy is used to filter target segments and prioritize writing metadata to hot segments to reduce the silent time.

Benefits of technology

It enables rapid recovery of hot-segmented data, reducing fault recovery time from minutes to seconds, and improving the availability and performance of the storage system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122111742B_ABST
    Figure CN122111742B_ABST
Patent Text Reader

Abstract

The application discloses a fault recovery method and an electronic device, and relates to the technical field of storage. The method comprises the following steps: receiving user read-write request data, judging whether the metadata corresponding to the block segment where the user read-write request data is located is in the cache pool; in response to the metadata corresponding to the block segment where the user read-write request data is located not being in the cache pool, screening the cache pool by using a hot and cold data elimination strategy, and determining a target block segment according to the hot and cold priorities of a plurality of block segments; performing disk flushing on the metadata corresponding to the target block segment; and in response to the disk flushing of the metadata corresponding to the target block segment being completed, recovering the controller node fault by processing the user read-write request data in the target block segment waiting queue. The application uses the hot and cold management mode of data area segmentation statistics, adjusts the priority level of segment metadata writing, and ensures that the hot block segment recovers the business as soon as possible.
Need to check novelty before this filing date? Find Prior Art