Copy-on-Demand Optimization for Large Writes in Data Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data storage systems face inefficiencies during protected restore operations due to unnecessary data copying and resource wastage, especially with large block sizes and multi-file blocks, leading to unprotected restores and resource inefficiencies.
Innovation Solution
A method and system for copy-on-demand optimization that determines if a target portion of the source data storage entity is synchronized with the destination, and if not, performs a selective copy-on-demand operation only for blocks that will be completely overwritten, reducing unnecessary data copying and resource usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a conventional copy-on-demand operation is performed during protected restore, then data is synchronized from destination to source, but unnecessary data copying occurs for blocks that will be completely overwritten, wasting resources and time
Solution Approach 1:
The patent applies partial action by performing copy-on-demand only for the specific subset of blocks that will not be completely overwritten by the pending large write. Instead of copying all blocks in the target portion, the system identifies and copies only those blocks that need to be synchronized, avoiding the excessive copying of blocks that will be overwritten anyway. This resolves the contradiction by making the copy operation partial rather than complete.
Solution Approach 2:
The patent segments the target portion into individual blocks and evaluates each block independently to determine whether it will be completely overwritten by the large write. This segmentation allows the system to selectively copy only the necessary blocks rather than treating the entire target portion as a single unit, thereby improving copy operation efficiency while maintaining data synchronization reliability.
2Reliability
If copy-on-demand is performed for all blocks in target portion, then data integrity is maintained, but resource consumption and power usage increase unnecessarily
Solution Approach 1:
The patent performs copy-on-demand only for blocks that will not be completely overwritten, applying partial action to avoid the energy waste of copying blocks that are about to be overwritten anyway. This maintains data integrity for the blocks that need synchronization while significantly reducing power consumption and resource usage compared to copying all blocks in the target portion.
3Reliability
If copy-on-demand is performed for all blocks in target portion, then data synchronization is ensured, but operation time is excessive due to unnecessary copying
Solution Approach 1:
The patent reduces copy operation time by performing copy-on-demand only for blocks that will not be completely overwritten by the large write. By identifying and excluding blocks that will be overwritten from the copy operation, the system maintains data synchronization reliability for necessary blocks while significantly reducing the time required compared to copying all blocks in the target portion.
Solution Approach 2:
The patent segments the target portion into individual blocks and evaluates each block's overwrite status independently. This segmentation enables the system to identify and exclude overwriteable blocks from the copy operation, thereby reducing operation time while ensuring data synchronization is maintained for blocks that actually need copying.
4Productivity
If conventional copy-on-demand is used during protected restore with large writes, then restore operation completes, but resource wastage and wear increase
Solution Approach 1:
The patent applies partial action by copying only the blocks that will not be completely overwritten by the large write during protected restore. This ensures the restore operation completes successfully with proper data synchronization while avoiding the resource wastage of copying blocks that are about to be overwritten, thereby reducing both material and energy resources consumed.
Data Source
AI summary
Methods, systems, and computer readable media for copy-on-demand optimization for large writes are disclosed. According to one aspect, a method for copy on demand optimization for large writes includes, in a system having a source data storage entity and a destination data storage entity, receiving a write request to write data to a target portion of the source data storage entity, the target portion of the source data storage entity comprising at least one data storage block. The method includes determining whether the source data storage entity and the destination data storage entity are engaged in a backup operation in which data sufficient to synchronize the source data storage entity to the destination data storage entity is copied from the destination data storage entity to the source data storage entity, and if so, determining whether the target portion of the source data storage entity is not yet synchronized to the destination data storage entity. If the target portion of the source data storage entity is not yet synchronized to the destination data storage entity, a copy-on-demand operation is performed to synchronize the target portion of the source data storage entity to the destination data storage entity, where each data storage block within the target portion of the source data storage entity that will be completely overwritten as a result of the write request is not synchronized to the destination data storage entity. The write request to write data to the target portion of the source data storage entity is then processed.