Data Shadowing System for Point-in-Time Restoration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data backup systems are laborious and confusing for users, fail to efficiently track and store the state of multiple file systems over time, and struggle with correct disk-level and file-level restoration to a point-in-time without storing redundant data.
Innovation Solution
The Data Shadowing System, which includes a memory module connected to the computer system via a communication medium, automatically stores data in a single format, using a control software component to maintain records and control operations, efficiently tracking file system changes and enabling point-in-time restoration without redundant data storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional data backup systems are used to store copies of data, then data can be restored after loss, but the storage requirements are considerable and the process is laborious and confusing
Solution Approach 1:
The shadowing driver automatically performs backup operations without user intervention. The system self-manages the shadowing process, tracking file system changes and storing shadow copies autonomously, eliminating the need for users to configure or manage backup tasks manually
Solution Approach 2:
A shadowing driver acts as an intermediary between the file system and storage device. This driver intercepts file system calls and automatically creates shadow copies of changed data blocks, transparently managing backup operations without requiring user knowledge or action
2Measurement precision
If change tracking is performed to monitor file system modifications, then data changes can be captured for backup, but memory usage increases due to redundant storage
Solution Approach 1:
The system extracts only the essential shadowing information needed for backup - specifically, the identities of changed data blocks and their new locations. Rather than storing complete file copies or extensive metadata, only the minimal necessary information is retained in shadowing structures
Solution Approach 2:
Shadow copies are created locally within the storage structure rather than requiring separate backup storage. The system maintains shadowing information in-place using data block pointers and shadowing directories that integrate with the existing file system structure, eliminating redundant storage
3Reliability
If complete data copies are stored for backup purposes, then data can be restored, but storage space is wasted and organization becomes complicated
Solution Approach 1:
Instead of copying entire files or data sets, the system creates shadow copies of only the specific data blocks that have changed. This selective copying approach maintains backup capability while dramatically reducing storage requirements and simplifying data organization
Solution Approach 2:
The backup system is segmented into discrete data block-level operations rather than file-level or volume-level operations. Each changed data block is independently tracked and shadowed, allowing for granular backup management and simplified organization of backup data
Data Source
AI summary
The Data Shadowing System comprises a memory module that is connected to the monitored computer system via an existing input/output port to store the shadowed data. The memory module includes a memory device for data storage as well as software, including a control software component that is automatically installed on the monitored computer system when the memory module is first connected to the monitored computer system, as well as associated module software for maintaining a record of the data stored on the memory device. The Data Shadowing System automatically stores the data on the memory module in a single format, while representing it in a data management database in two formats: disk sectors and files. The Data Shadowing System thereby efficiently tracks and stores the state of multiple file systems over time, while allowing for correct disk-level and file-level restoration to a point-in-time without storing redundant data.


