Distributed Data Protection via Copy-on-Write and Parity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional data protection solutions are costly to maintain and implement due to administrative and infrastructure expenses, and they do not efficiently manage the growing volume of data, especially in scenarios where data loss occurs due to hardware failures or natural disasters.
Innovation Solution
A distributed data protection system where a data protection program identifies and tracks duplicate and unique data objects across network hosts, breaks unique data objects into chunks, and distributes them across the network, maintaining parity to enable quick recovery in case of host failures, reducing the need for expensive backup infrastructure and administrative overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional backup solutions are implemented, then data protection is achieved, but administrative and infrastructure costs increase significantly
Solution Approach 1:
The patent combines backup functionality directly into the file system layer, merging data protection with the operating system's native file management. This eliminates the need for separate backup hardware and software infrastructure, reducing costs while maintaining data protection through copy-on-write operations that occur transparently during normal file operations.
Solution Approach 2:
The system performs automatic data protection through copy-on-write operations that occur transparently during normal file operations. The backup mechanism serves itself by utilizing the file system's existing structure and operations, eliminating the need for external backup software and reducing administrative overhead while maintaining continuous data protection.
2Ease of operation
If data is stored on network hosts, then data availability is maintained, but data loss risk increases due to hardware failures or natural disasters
Solution Approach 1:
The system performs preliminary backup actions through copy-on-write operations that create backup copies of data blocks before the original data can be lost. This preliminary protection ensures that even if hardware fails or natural disasters occur, the backup copies are already in place and can be used for recovery, eliminating data loss risk while maintaining data availability.
Solution Approach 2:
The system establishes protective cushioning by creating backup copies of data blocks in advance through copy-on-write operations. This beforehand cushioning ensures that when hardware failures or natural disasters occur, the backup copies serve as a protective buffer, preventing data loss while maintaining continuous data availability on network hosts.
3Quantity of substance
If data is divided into chunks and distributed across network hosts, then storage efficiency improves, but system complexity increases
Solution Approach 1:
The patent segments data into individual blocks that can be independently managed and distributed across network hosts. Each data block is treated as a separate unit that can be backed up, stored, and recovered independently. This segmentation improves storage efficiency by eliminating redundant backups of identical data blocks while the file system layer manages the complexity of distribution transparently.
Solution Approach 2:
The file system layer acts as an intermediary that manages the complexity of data chunking, distribution, and recovery. It translates high-level file operations into low-level block operations, handling the segmentation and distribution of data across network hosts while presenting a simple, unified interface to users and applications, thus improving storage efficiency without exposing system complexity.
4Reliability
If backup storage devices are purchased, then data protection capacity increases, but power costs and maintenance expenses increase
Solution Approach 1:
The patent makes existing network hosts and file systems perform multiple functions: they simultaneously handle normal file operations and data protection/backup operations. This multi-functionality eliminates the need for dedicated backup storage devices, reducing power costs and maintenance expenses while increasing data protection capacity by utilizing the computational resources already present in the network infrastructure.
Solution Approach 2:
The system uses the network hosts' own computational resources to perform backup operations, making them self-service rather than requiring external dedicated backup hardware. The file system on each host manages its own data protection, utilizing existing processors, storage, and network connections, thereby eliminating power costs and maintenance expenses associated with separate backup infrastructure while expanding data protection capacity.
Data Source
AI summary
Data protection programs are installed at each network host. The programs communicate with each other to scan the hosts and identify duplicate and unique data objects stored at the hosts. Duplicate data objects are maintained on the hosts. Unique data objects are broken into chunks, copied to other hosts, and a parity data is calculated. When a network host becomes unavailable and is replaced with a new network host, duplicate data objects stored on the now unavailable network host may be rebuilt on the new network host using the maintained duplicate data objects on the other hosts. Unique data objects stored on the now unavailable network host may be rebuilt on the new network host using the copied chunks and parity data.