Secondary Host Off-Host Backup Using Physical Block Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing file system backup methods are inefficient as they require significant processing resources from the primary host to create backups of clone file sets, which share data with active file sets, and do not effectively distribute the processing load to secondary hosts.
Innovation Solution
A method where a secondary host identifies contiguous storage locations on a storage device to read data sequentially without accessing the file system, using physical location information from the file system to construct a copy of the clone file set and write it to backup storage, thereby offloading processing resources from the primary host.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If backup software writes files to backup storage on a file-by-file basis reading all data from storage device, then backup completeness is ensured, but backup time increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-identifying and cataloging all data block locations for files before the backup process begins. The mapping catalog stores the physical locations of all data blocks, allowing the backup system to retrieve multiple files simultaneously in parallel operations rather than sequentially processing files one at a time, thus reducing total backup time while maintaining completeness
Solution Approach 2:
The patent segments the backup process into independent parallel operations by identifying non-contiguous data blocks for multiple files and retrieving them simultaneously. Each data block or chunk can be processed independently and in parallel, transforming a sequential file-by-file backup into a concurrent block-level backup operation that significantly reduces backup time
2Quantity of substance
If clone file set shares data blocks with primary file set on same device, then storage efficiency is improved, but backup processing resources on host increase
Solution Approach 1:
The patent creates a mapping catalog that is a copy of the file system's metadata structure, storing information about data block locations. This catalog allows the backup system to access and process clone file set data without requiring the primary host's processing resources, as the mapping information can be read directly from storage devices by secondary backup systems
Solution Approach 2:
The patent extracts the backup processing function from the primary host system by enabling secondary hosts to directly access storage devices and read data using the mapping catalog. This extraction removes the backup processing load from the primary host, allowing it to focus on maintaining the primary file set while backup operations are performed independently on separate systems
3Productivity
If secondary host reads data sequentially from contiguous storage locations, then backup speed increases, but ability to handle non-contiguous file data decreases
Solution Approach 1:
The patent introduces a mapping catalog as an intermediary between the file system and the backup process. This catalog contains information about the physical locations of all data blocks, allowing the backup system to efficiently navigate and retrieve non-contiguous data blocks. The mapping catalog enables secondary hosts to read data in an optimized sequence while still handling files with scattered blocks across the storage device
Data Source
AI summary
A method, system, computer system, and computer-readable medium that enable a secondary host that is not the file system host to create a backup of a clone file set that shares at least one data block on a storage device with an active file set. Start and end locations are identified for a set of contiguous storage locations (referred to as a “chunk”) on the storage device. Physical location information is obtained for each portion of a file contained in the chunk. The start and end locations and physical location information for portions of files contained in the chunk are provided to the secondary host, which sequentially reads data from the set of contiguous storage locations and constructs a copy of the file(s) making up the clone file set. The file(s) are written by the secondary host to a storage device to create a backup of the clone file set.