Dynamic Quorum Disk Selection in Server Clusters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional split-brain solutions in distributed file storage systems require dedicated quorum disks, leading to increased costs and administrative overhead due to the need for extra resources and continuous monitoring, which are underutilized and idle.
Innovation Solution
Designating selected nodes in a server cluster as quorum disks, with a list of these nodes stored on each node, allowing for dynamic selection and update based on availability, eliminating the need for dedicated hardware and reducing administrative burden.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dedicated quorum disks are used for split-brain coordination, then communication reliability during network failures is improved, but system cost and administrative overhead increase
Solution Approach 1:
The patent applies universality by enabling regular network nodes to serve dual purposes: both data storage functions and quorum disk coordination functions. During normal operation, nodes perform data storage; during split-brain scenarios, the same nodes execute quorum coordination through versioned list comparison, eliminating the need for dedicated quorum hardware.
Solution Approach 2:
The patent merges the quorum coordination function with existing network nodes by integrating versioned list storage and comparison capabilities into regular node operations. This combines previously separate functions (data storage and quorum coordination) into a unified system using the same hardware resources.
2Reliability
If dedicated quorum disks are continuously monitored to ensure availability, then data integrity is improved, but administrative overhead increases
Solution Approach 1:
The patent implements self-service through automated versioned list management. Nodes automatically generate, store, and compare versioned quorum lists without administrator intervention. The system self-monitors availability by attempting list retrieval and automatically handles failures through version comparison and fallback mechanisms.
Solution Approach 2:
The patent uses feedback mechanisms where nodes continuously attempt to retrieve and compare quorum lists, receiving feedback about list availability and version consistency. This feedback drives automatic adjustment of quorum participation and triggers appropriate error handling when inconsistencies are detected.
3Reliability
If the number of quorum disks is increased to improve fault tolerance, then reliability during split-brain events is improved, but resource utilization decreases
Solution Approach 1:
The patent applies dynamics by making the quorum disk set dynamic rather than static. The system automatically adjusts which nodes serve as quorum disks based on current cluster state, availability, and versioned list consistency. Nodes can dynamically enter or exit quorum participation without manual reconfiguration, optimizing resource utilization while maintaining fault tolerance.
Data Source
AI summary
A node in a server cluster is designated as a quorum disk. The node stores a list of other nodes in the server cluster also designated as quorum disks. The node can replace the first list with a second and more recent list of quorum disks only if the second list is updated on at least a simple majority of quorum disks on the first list.