Distributed Data Authority System for Eventual Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data management systems that rely on a single master hierarchy to maintain data consistency across multiple storage sites are time-intensive and expensive, particularly when strong consistency is required, as they necessitate synchronization of all duplicate data files before changes can be made to the master data file.
Innovation Solution
A distributed data authority system that utilizes a replication set with an edit authority object, allowing only one data storage site to have edit authority at a time, which is transferable, and an ownership authority object to manage which sites can receive edit authority, enabling eventual consistency across sites without requiring instantaneous synchronization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single master hierarchy is used to maintain data consistency, then data consistency is ensured, but the time and expense for maintaining consistency increases
Solution Approach 1:
The patent segments the centralized master hierarchy into multiple distributed data storage sites, each capable of independently storing and managing data objects. This segmentation eliminates the single point of control, allowing parallel operations across sites while maintaining consistency through version vectors that track changes independently at each location.
Solution Approach 2:
The patent introduces version vectors as an intermediary mechanism that mediates consistency across distributed sites. Version vectors automatically track and compare changes without requiring centralized coordination, enabling the system to detect and resolve conflicts without the overhead of master hierarchy synchronization.
2Reliability
If strong consistency is required, then data accuracy is maintained, but the cost and time for synchronization increases
Solution Approach 1:
The patent applies partial action by implementing consistency only where and when needed through version vector comparison. Instead of forcing full synchronization across all sites for every operation, the system performs minimal consistency checks only for conflicting updates, reducing the overhead while maintaining accuracy for critical data.
Solution Approach 2:
The patent changes the parameter of consistency from a binary state (consistent/inconsistent) to a graded state using version vectors that capture the degree and timing of changes. This allows the system to determine consistency based on version comparisons rather than requiring complete synchronization, improving productivity while maintaining data accuracy.
3Reliability
If all duplicate data files are synchronized before changes, then consistency is maintained, but the time-intensive process increases expense
Solution Approach 1:
The patent performs preliminary action by pre-configuring version vectors with each data object, establishing a baseline for tracking changes. This preliminary setup eliminates the need for repeated full synchronization processes, as the version vectors are already in place to automatically track and compare updates without requiring energy-intensive coordination.
Solution Approach 2:
The patent uses version vectors as lightweight copies of change metadata that can be replicated and compared without copying the actual data content. This allows the system to track consistency through small, inexpensive metadata structures rather than requiring expensive full data synchronization operations.
Data Source
AI summary
An apparatus, system, and method for distributing authority for data objects. Each of the data objects in a replication set is stored at a different data storage site and comprises a copy of a content. An edit authority is assigned to only one of the data storage sites at a time using an edit authority object corresponding to the replication set that is transferable between the data storage sites. Responsive to a change to the content of a data object located at a particular data storage site having the edit authority, an update is propagated to a remaining portion of the data storage sites to affect eventual consistency of the replication set. Usability of the data objects at the remaining portion of the data storage sites is maintained until the update is received and processed to thereby reduce a time and expense associated with maintaining data consistency.