Distributed Storage Journal Segmentation for Network Load Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional distributed storage systems face high network loads during data redundancy operations, which can bottleneck performance and are inefficient in managing write histories for both original and redundant data.
Innovation Solution
A distributed storage system architecture that includes multiple physical storage devices and computers connected via a network, where write requests are processed by distributing write data and redundant data across these devices, with collective journal management to reduce network load and ensure data integrity through guarantee codes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is redundantly stored in a plurality of storage devices, then data reliability is improved, but network load increases
Solution Approach 1:
The patent segments the journal writing process into two distinct parts: a first journal for original data and a second journal for redundant data. This segmentation allows each journal to be managed independently, enabling parallel processing and reducing the overall network load while maintaining data reliability through redundant storage.
Solution Approach 2:
The patent implements preliminary action by writing the first journal for original data before writing the second journal for redundant data. This preliminary action ensures that the original data is recorded first, establishing a foundation for subsequent redundant data writing, and allows for optimized resource allocation during the redundancy process.
2Reliability
If journals for original data and redundant data are written separately, then data integrity is improved, but processing time increases
Solution Approach 1:
The patent maintains continuity of useful action by implementing an overlapping execution model where the writing of the first journal and the writing of the second journal are performed in an overlapping manner. This allows both journals to be processed continuously without idle time, ensuring data integrity through separate journal management while minimizing processing time through parallel execution.
Solution Approach 2:
By performing the first journal writing as a preliminary action before the second journal writing, the system establishes a structured sequence that ensures data integrity. The preliminary first journal captures original data state, and the subsequent second journal captures redundant data state, with both processes optimized to run efficiently in sequence and overlap where possible.
Data Source
AI summary
A distributed storage system includes one or a plurality of storage units including a plurality of physical storage devices, and a plurality of computers connected to the one or plurality of storage units via a communication network. When receiving a write request for a logical volume, the computer writes write data corresponding to the write request and redundant data for making the write data redundant in a plurality of physical storage devices of the storage unit in a distributed manner, and collectively controls writing of a journal of write data for managing a write history of the write data and a journal of redundant data for managing a write history of the redundant data.