Delta Data Backup Using Reference Blocks to Cut Sync Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional asynchronous backup functions require large data transmission and computation, leading to high resource usage and costs, and are inefficient for low-end destination storage systems due to lack of inline data de-duplication functionality and lengthy synchronization times.
Innovation Solution
A method that identifies frequently-used delta data blocks and transmits only delta index and reference information, along with additional data blocks, to reduce data transmission and computation, optimizing backup strategies based on user requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional asynchronous backup functions transmit all delta data, then data backup completeness is ensured, but data transmission volume and computation resources are greatly increased
Solution Approach 1:
The patent extracts and identifies duplicate data blocks within the delta data using de-duplication technology. Only the unique (non-duplicate) data blocks are transmitted to the backup storage system, while duplicate blocks are referenced from the previous data. This extraction of unnecessary duplicate transmissions directly reduces data transmission volume while maintaining backup completeness through reference mechanisms.
Solution Approach 2:
The patent changes the transmission parameter from transmitting all delta data blocks to transmitting only unique data blocks. By introducing de-duplication identification and reference mechanisms, the system transforms the backup process from full delta transmission to selective unique block transmission, significantly reducing the quantity of transmitted data while ensuring completeness through reference pointers.
2Reliability
If traditional asynchronous backup functions transmit all delta data, then data backup completeness is ensured, but processing unit and memory resources are greatly used
Solution Approach 1:
The patent extracts duplicate data blocks through de-duplication identification, separating unique blocks from duplicates. This extraction reduces the computational burden on processing units by avoiding redundant processing of duplicate blocks and reduces memory usage by not storing redundant copies in the backup system, while maintaining completeness through reference mechanisms.
Solution Approach 2:
Instead of physically copying and transmitting duplicate data blocks, the patent uses reference pointers to point to the original blocks in the previous data. This virtual copying approach maintains data completeness and accessibility while dramatically reducing the actual data transmission and storage resources required.
3Ease of manufacture
If low-end destination storage systems are used without inline data de-duplication functionality, then cost is reduced, but synchronization time becomes lengthy
Solution Approach 1:
The patent performs de-duplication identification and unique block determination at the source storage system before transmission. This preliminary processing at the source eliminates the need for complex inline de-duplication functionality at the destination system, allowing low-end cost-effective systems to be used while maintaining efficient synchronization speeds through pre-processed unique block identification.
Solution Approach 2:
The patent introduces a reference mechanism as an intermediary between the source and backup systems. The reference information acts as a mediator that allows the backup system to efficiently identify and handle duplicate blocks without requiring expensive inline de-duplication hardware or software at the destination, thus reducing cost while maintaining synchronization efficiency.
Data Source
AI summary
Embodiments of the present disclosure relate to a data backup method, a data backup device, and a computer program product. The method includes: determining delta data based on previous data and current data of a storage system; determining a delta data block subset in a delta data block set; sending delta index information and delta reference information associated with delta data blocks in the delta data block subset to a backup storage system; and sending, to the backup storage system, a further delta data block subset including delta data blocks in the delta data block set other than the delta data block subset. With the technical solution of the present disclosure, the amount of data transmission, the amount of computation, and the usage of a processing unit and a memory when backing up data can be reduced.