Data Replica Change Analysis Using Identifier-Based Segment Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for approximating un-replicated data using time-based measurements are inaccurate and do not account for transmission errors or storage failures, and determining which replicated data is present at which destination machine is time-consuming and error-prone.
Innovation Solution
A process that receives an asynchronous data update with a stream of identifiers and a separate stream of data, using in-sync and out-of-sync data structures to track data segments, allowing direct measurement of un-replicated data and identifying which data is present at which destination machine.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If time-based measurements are used to approximate un-replicated data, then the measurement process is simple, but the accuracy is poor and does not account for transmission errors or storage failures
Solution Approach 1:
The patent segments data into discrete data segments with unique identifiers, allowing individual tracking of each segment's replication status. This segmentation enables precise measurement of un-replicated data by counting specific segments rather than using imprecise time-based approximations, directly resolving the accuracy problem while maintaining manageable system complexity through structured data organization
Solution Approach 2:
The patent implements feedback mechanisms where the replication system continuously monitors and reports the actual replication status of data segments. This feedback loop allows the system to account for transmission errors and storage failures by detecting when segments fail to replicate and adjusting the measurement accordingly, significantly improving measurement precision without requiring overly complex infrastructure
2Productivity
If traditional methods are used to determine which replicated data is present at which destination machine, then the process is simpler, but it is time-consuming and error-prone
Solution Approach 1:
The patent performs preliminary actions by pre-establishing a mapping between data segment identifiers and their replication status before actual replication occurs. This preliminary setup includes creating data structures that track expected versus actual replicated segments, enabling rapid determination of data presence at destination machines without time-consuming verification processes, thus improving both productivity and reliability
Solution Approach 2:
The patent uses copying of data segment identifiers and status information to track replication state across multiple destination machines. Instead of directly querying each destination machine for data presence (which is time-consuming and error-prone), the system maintains copies of the replication state information, allowing fast and reliable determination of which data is present where through these replicated status records
Data Source
AI summary
An embodiment enables a first machine to receive an asynchronous data update from a second machine, the asynchronous data update comprising a first stream of identifiers and a second stream of data, wherein the first stream of identifiers comprises a first identifier corresponding to a first data segment in the second stream of data. An embodiment analyzes, as a part of the asynchronous data update, the first stream of identifiers. An embodiment determines at a first time, responsive to the analyzing, that the first data segment corresponding to the first identifier has not arrived at the first machine by the first time. An embodiment updates, responsive to the determining at the first time, a record at the first machine with the first identifier. An embodiment updates the record at a second time, responsive to receiving the first data segment at the first machine at the second time.