Dynamic Data Protection Based on Change Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data protection schemes are inadequate as they do not account for the operational significance of data and often waste resources by protecting entire data sets even when only a small portion has changed, and they fail to provide timely protection against data corruption and replication of infected data.
Innovation Solution
A data protection system that dynamically triggers protection mechanisms based on the type and class of data and the amount of data changed since a prior point in time, allowing for tailored protection resources allocation and real-time updates, including incremental backups, continuous replication, and virus scanning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If time-based scheduled protection is used, then protection operations are performed at regular intervals, but large amounts of unprotected data exist between protection operations and resources are wasted protecting entire data sets even when only small portions have changed
Solution Approach 1:
The system dynamically adjusts protection operations based on real-time data change detection. Instead of fixed scheduled protection, the system continuously monitors data modifications and triggers protection operations only when changes are detected, making the protection mechanism adaptive to actual data states and eliminating wasted resources on unchanged data.
Solution Approach 2:
The protection system automatically detects data changes and initiates protection operations without external scheduling. The system monitors its own data environment, identifies when protection is needed, and executes appropriate protection actions autonomously, eliminating the need for manual scheduling while ensuring timely protection.
2Reliability
If entire data sets are protected with each backup operation, then complete data protection is ensured, but significant resources are wasted when only a small portion of data has changed
Solution Approach 1:
The system segments the data set into individual files or data units and monitors changes at this granular level. When a change is detected, only the specific changed segment is protected rather than the entire data set. This segmentation enables precise protection of only what is necessary, dramatically improving efficiency while maintaining complete protection of all changed data.
Solution Approach 2:
The protection mechanism applies different protection actions to different parts of the data set based on their change status. Changed data receives protection operations while unchanged data remains untouched. This local quality approach ensures that protection resources are concentrated where needed rather than uniformly applied across all data.
3Reliability
If mirrored copies are maintained locally, then data availability is improved, but security breaches can corrupt all accessible resources including mirrored copies
Solution Approach 1:
The system introduces an intermediary protection layer between the data and potential corruption sources. Virtual machine snapshots and isolation mechanisms act as intermediaries that prevent direct corruption of mirrored copies even when security breaches occur. The intermediary layer captures and contains corruption attempts, protecting the underlying data and mirrors from direct impact.
Solution Approach 2:
The system creates protective snapshots of data before potential corruption events can occur. These snapshots serve as cushioning backup states that can be restored if corruption is detected. By preparing these protective measures in advance, the system ensures that even if mirrors are compromised, the original data can be recovered from the cushioning snapshots.
Data Source
AI summary
A method, system, computer system, and computer-readable medium to trigger protection of a set of data based upon the type or class of the data in the set and/or an amount of data that has changed since some prior point in time. Types of protection that can be triggered include full backup, incremental backup, switching to a different type of continuous replication, intermittent replication, and virus detection. Using the type of the data as a basis for triggering protection enables the operational significance of different sets of data to be taken into account when allocating protection resources. Data sets may be pre-classified, or an automated determination of the type of a data set may be determined by variables measured at run-time. The amount of the changed data in a set that triggers protection may vary in accordance with the type of the set of data.