Deduplicated Data Extent Restoration via Sequential Sorting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for restoring deduplicated data from sequential backup devices suffer from poor access performance due to the random access nature of disk-based storage, leading to inefficient retrieval and transmission of deduplicated data objects.
Innovation Solution
The backup server sorts and retrieves deduplicated data extents from sequential backup devices in the order they were stored, allowing the client to reconstruct data objects efficiently by accessing the extents sequentially, optimizing storage and network bandwidth usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If deduplicated data is stored on disk with random access capability, then data retrieval flexibility is improved, but restoration performance deteriorates due to inefficient sequential access patterns
Solution Approach 1:
The patent pre-sorts deduplicated extents by their physical location on sequential backup devices before restoration operations. This preliminary organization creates an access map that enables sequential retrieval patterns, transforming random access requirements into efficient sequential access and thereby improving restoration performance without sacrificing retrieval flexibility
Solution Approach 2:
The patent introduces an intermediary sorting layer between the disk-based deduplicated storage and the sequential backup devices. This intermediary structure (the sort list mapping extents to physical locations) acts as a mediator that translates random access patterns from disk into sequential access patterns for tape, resolving the contradiction between access flexibility and restoration efficiency
2Reliability
If deduplicated extents are transmitted from server to client for multiple data objects, then complete data reconstruction is achieved, but network bandwidth usage increases due to redundant data transmission
Solution Approach 1:
The patent extracts and transmits only the unique deduplicated extents that are not already present on the client side. By identifying and removing redundant extents before transmission, the system achieves complete data reconstruction while minimizing network bandwidth consumption, as clients receive only the necessary unique data blocks
Solution Approach 2:
The patent implements a mechanism where the server identifies extents that can be discarded (already exist on client) and those that must be recovered (missing on client). This selective transmission approach ensures data completeness while optimizing network usage by discarding redundant transmissions and only recovering necessary extents
3Reliability
If multiple volume mount and seek operations are performed on sequential backup devices, then complete data retrieval is achieved, but access performance deteriorates
Solution Approach 1:
The patent performs preliminary sorting of extents by their physical location on sequential backup devices before the restoration process begins. This pre-organization ensures that during actual data retrieval, the system can access extents in sequential order without requiring multiple volume mounts or seek operations, thereby maintaining both completeness and speed
Solution Approach 2:
The patent enables continuous sequential access to deduplicated extents on tape by organizing them in physical storage order. This continuity eliminates interruptions caused by volume mounting and seeking operations, allowing the restoration process to proceed with uninterrupted sequential reading, thus improving access performance while ensuring complete data retrieval
Data Source
AI summary
Provided are computer program product, system, and method for restoring deduplicated data objects from sequential backup devices. A server stores data objects of extents having deduplicated data in the at least one sequential backup device. The server receives from a client a request for data objects. The server determines extents stored in the at least one sequential backup device for the requested data objects. The server or client sorts the extents according to an order in which they are stored in the at least one sequential backup device to generate a sort list. The server retrieves the extents from the at least one sequential backup device according to the order in the sort list to access the extents sequentially from the sequential backup device in the order in which they were stored. The server returns the retrieved extents to the client and the client reconstructs the requested data objects from the received extents.