Data Restoration Using Forward and Backward Deltas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional online backup systems face performance issues and increased operational demands when restoring recent policies, as they maintain the baseline at the end of the retention period, requiring frequent rotation through versions, leading to high costs and power consumption for storage and cooling.
Innovation Solution
The system maintains a baseline closer to the present time and stores both newer and older versions, allowing for faster restoration by combining fewer versions with the baseline, and relocates older versions to lower-cost, low-power storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If the baseline is maintained at the end of the retention period (oldest data), then storage capacity is optimized, but restoration performance for recent policies deteriorates due to requiring rotation through most or all versions
Solution Approach 1:
The patent segments the version chain into two parts: a baseline containing recent versions for fast restoration, and archived older versions stored separately. This segmentation allows recent policies to be restored quickly using only the baseline, while older versions are available when needed without impacting recent restoration performance.
Solution Approach 2:
The patent extracts older versions from the main baseline and stores them in separate archive storage. This extraction reduces the size of the active baseline, enabling faster restoration operations for recent policies while preserving access to historical data through the archived versions.
2Productivity
If all versions are kept on online storage or high-end near-line storage to provide fast restoration, then restoration performance improves, but cost and power consumption increase
Solution Approach 1:
The patent applies local quality by providing high-performance storage only where needed (the baseline containing recent versions) and using lower-performance, lower-cost storage for older versions. This creates a tiered storage architecture where each layer provides appropriate performance for its specific use case, optimizing both speed and energy consumption.
Solution Approach 2:
The patent implements a dynamic baseline rotation mechanism where the baseline automatically adjusts to contain only the most recent versions needed for fast restoration. As versions age, they are automatically archived to slower storage, allowing the system to dynamically optimize the balance between restoration speed and storage resource utilization.
3Reliability
If the baseline is rotated forward continuously to maintain retention period compliance, then data retention requirements are met, but restoration operations for recent policies require more steps
Solution Approach 1:
The patent performs preliminary action by pre-archiving older versions before they are needed for restoration. The baseline rotation process proactively moves versions to archive storage in advance, ensuring that the baseline remains compact and optimized for fast recent restorations while compliance with retention periods is maintained through the archived copies.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A restoration technique includes storing a collection of blocks having (i) a first baseline of blocks defining a file at a first time within a predefined window, (ii) older blocks, and (iii) newer blocks. The technique further includes providing a direction signal to restore the file at a particular time from the collection of blocks. The direction signal has an older direction value when the particular time is older than the first time and (ii) a newer direction value when the particular time is newer than the first time. The technique further includes forming, from the collection of blocks, a second baseline of blocks from the first baseline of blocks and additional blocks. The additional blocks are either (i) the older blocks or (ii) the newer blocks depending on the direction value. The second baseline defines the file at the particular time.