Distributed File System Metadata Backup via Dormant Servers

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In distributed file systems, existing technologies face a single point of failure when remote disaster tolerance is implemented, as metadata and actual data backups are not effectively managed, leading to potential loss during disasters.

Innovation Solution

A distributed file system architecture with dormant metadata servers and alternate file access servers at both main and backup locations, utilizing heartbeat detection and periodic synchronization to ensure metadata and data backups, enabling seamless switching and dual-copy storage strategies to prevent data loss.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional backup methods are used to back up metadata and actual data at a remote location, then data backup is achieved, but a single point of failure exists when disaster occurs at the backup location

Engineering Contradiction:
Improvedisaster toleranceVSAvoidbackup architecture
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system segments the backup architecture into multiple independent components: main FLR, backup FLR, dormant FLRs, and alternate FASs. Each component has a specific role and can operate independently, eliminating the single point of failure by distributing functionality across multiple segmented units rather than relying on a single backup location.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements preliminary action by pre-configuring dormant FLRs and alternate FASs at backup locations before disasters occur. These components are prepared in advance with the necessary software and configuration, so when a disaster strikes, they can immediately activate and take over without requiring complex post-disaster setup or recovery procedures.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If only one copy of metadata and actual data is stored at backup location, then storage space is optimized, but data loss occurs when failure happens at backup location

Engineering Contradiction:
Improvedata securityVSAvoiddata copies
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The system applies local quality by creating different copies of data with different purposes and locations. Multiple copies of metadata are stored at the backup location, and additional alternate FASs store copies of actual data. Each copy serves a specific local function, ensuring that if one copy is lost, other copies remain intact and can be used for recovery.

Inventive Principle:
Principle #3Local quality

3Productivity

If rapid switching is implemented from main FLR to backup FLR, then service continuity is improved, but the recovered system still has single point of failure

Engineering Contradiction:
Improveservice continuityVSAvoidsystem robustness
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system implements dynamics by enabling flexible role transitions between components. Dormant FLRs can dynamically switch to become active backup FLRs, and alternate FASs can dynamically take over from primary FASs. This dynamic capability allows the system to adapt to various failure scenarios and maintain service continuity while distributing risk across multiple components, eliminating the single point of failure.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP2902922B1Distributed file system and data backup method for distributed file system
Publication Date: 2017.02.22 ZTE CORP
  • EP2902922B1 patent drawingFigure 1
  • EP2902922B1 patent drawingFigure 2
  • EP2902922B1 patent drawingFigure 3~4

AI summary

Provided are a distributed file system and a data backup system for the distributed file system. The system includes: a main FLR, a first FAC, a main FAS, and at least one first dormant FLR and a first alternate FAS; a backup distributed subsystem comprises a backup FLR, a second FAC, a backup FAS, and at least one second dormant FLR and a second alternate FAS; the at least one first dormant FLR and the at least one second dormant FLR are both used to back up the metadata on the FLR or on the backup FLR; the first alternate FAS and the second alternate FAS are both used to synchronize with the main FAS and the backup FAS and to perform write operations on current real data when the first FAC or the second FAC receives data write operation commands. The solution enhances the reliability and practicality of the system.