Cloud Snapshot Redaction for Storage Overhead Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data protection systems face challenges in minimizing downtime during recovery and keeping pace with high data transaction rates, leading to potential production site shutdowns due to the overhead of journaling, which results in data backlog and limited recovery points in time.
Innovation Solution
The method involves creating snapshots at a replication site by determining differences between snapshots, creating redacted snapshots with pointers to common data and storing only differential changes, allowing for efficient data replication and recovery in object-based storage systems, including cloud storage, which enables instantaneous access and reduced storage overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If journaling is used to enable continuous data protection and rollback to any point in time, then the ability to recover to any specified point in time is improved, but the overhead of journaling causes the backup site to fall behind high data transaction rates at the production site, leading to production site shutdowns
Solution Approach 1:
The patent segments the continuous data protection functionality into discrete snapshots taken at specific intervals. Instead of maintaining continuous journaling overhead, the system divides protection into manageable snapshot units that can be independently managed and restored, reducing the ongoing overhead while preserving recovery capabilities.
Solution Approach 2:
The system implements periodic snapshots at defined intervals rather than continuous journaling. This periodic approach allows the backup site to catch up periodically without being overwhelmed by continuous high-volume transactions, maintaining recovery points while preventing production shutdowns.
2Reliability
If full snapshots are stored at every point in time to enable recovery, then recovery completeness is improved, but storage overhead increases significantly
Solution Approach 1:
The patent implements nested snapshots where later snapshots are contained within or reference earlier snapshots. Each snapshot stores only the differences or changes since the previous snapshot, creating a nested structure that minimizes redundant storage while maintaining the ability to restore to any point in the snapshot sequence.
Solution Approach 2:
The system discards redundant data in later snapshots by referencing common data in earlier snapshots. Instead of storing complete copies of all data at each snapshot point, the system keeps only the differential changes and can recover complete state by combining the base snapshot with subsequent differential snapshots.
3Reliability
If the backup site processes every data transaction in real-time to keep pace with the production site, then data synchronization is improved, but the backup site cannot handle high transaction rates without forcing production site shutdowns
Solution Approach 1:
The system performs preliminary actions by taking snapshots at predetermined intervals or triggers rather than processing every transaction in real-time. This allows the backup site to prepare and process data at manageable rates while still maintaining synchronization at critical points, preventing production shutdowns while ensuring data protection.
Solution Approach 2:
The patent allows the backup site to skip processing individual transactions between snapshots and instead capture the state at snapshot boundaries. This skipping approach enables the backup site to handle high transaction rates without being constrained by processing each transaction sequentially, maintaining synchronization at key moments without forcing production shutdowns.
Data Source
AI summary
A computer implemented method, system, and program product for creating snapshots instructions at a replication site, the method comprising determining the differences between a first snapshot at a first time of a LUN and a second snapshot at a second time of the LUN, determining the differences between the second snapshot and a third snapshot at a third time of the LUN, creating a redacted snapshot of the second snapshot wherein the redacted snapshot contains pointers to the first snapshot for the common data in the first snapshot and the second snapshot; wherein the redacted snapshot contains pointers to the third snapshot for common data in the second snapshot and the third snapshot; and further wherein the redacted snapshot stored data that is not present in the first snapshot and is not present in the third snapshot.