Distributed Node State Monitoring for Split Brain Prevention

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Distributed storage systems face challenges in quickly detecting node failures and preventing split brain states, leading to reduced availability and potential data inconsistency due to reliance on centralized control nodes or complex agreement processes.

Innovation Solution

Each information processing device in the system autonomously receives and shares state information with other devices, determining node states and transmitting this information to maintain a shared understanding of system health, thereby preventing split brain states and reducing reliance on centralized monitoring.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a centralized control node is used to monitor node states and detect failures, then the system can prevent split brain states, but the system availability decreases and the control node becomes a single point of failure

Engineering Contradiction:
Improveprevention of split brain stateVSAvoidsystem availability
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent divides the centralized monitoring function into distributed monitoring capabilities across multiple nodes. Each node independently monitors the system state and can detect failures without relying on a single control node, thereby eliminating the single point of failure while maintaining split brain prevention through coordinated state assessment among nodes

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent combines the monitoring functions of multiple nodes into a collaborative system where each node contributes to the overall state determination. By merging individual node perspectives through shared state information and coordination protocols, the system achieves both high availability and reliable failure detection without centralized control

Inventive Principle:
Principle #5Merging (Combining)

2Reliability

If a specific node or administrator performs failure determination instead of all nodes, then split brain states are prevented, but the time to detect and respond to failures increases

Engineering Contradiction:
Improveprevention of split brain stateVSAvoidfailure detection time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by having nodes continuously exchange state information and maintain updated views of system health before failures occur. This proactive state sharing enables rapid failure detection and response when issues arise, eliminating delays associated with centralized determination while preventing split brain states through pre-established coordination mechanisms

Inventive Principle:
Principle #10Preliminary action

3Productivity

If each node autonomously determines node states without centralized control, then system availability is improved, but split brain states may occur leading to data inconsistency

Engineering Contradiction:
Improvesystem availabilityVSAvoiddata consistency
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent implements feedback mechanisms where nodes continuously share their state determinations and assumptions about system health. This feedback loop allows autonomous nodes to coordinate their decisions, detect potential split brain conditions, and resolve inconsistencies while maintaining high availability through distributed decision-making rather than centralized control

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS10298478B2Information processing system, computer-readable recording medium having stored therein control program for information processing device, and control method of information processing system
Publication Date: 2019.05.21 FUJITSU LTD
  • US10298478B2 patent drawing
  • US10298478B2 patent drawing
  • US10298478B2 patent drawing

AI summary

An information processing system includes a plurality of information processing apparatuses connected to each other, and communication is performed between the plurality of information processing apparatuses. Each of the plurality of information processing apparatuses includes a processor. The processor included in one of the plurality of information processing devices serving as a first information processing device receives, from each of second information processing devices other than the first information processing device, state information regarding each state of the plurality of information processing devices determined by the second information processing devices, determines each state of the plurality of information processing devices, based on the state information received from each of the second information processing devices, and transmits state information for transmission regarding the determined state of each of the plurality of information processing devices to each of the second information processing devices.