Cloud Storage Data Reduction With Centralized Write Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems face inefficiencies in data management and reliability, particularly in cloud-based environments, due to redundant write operations and lack of centralized control over storage processes, leading to increased latency and reduced performance.
Innovation Solution
Implementing a direct-mapped flash storage system with centralized control by the operating system, which directly addresses data blocks without translation by storage controllers, and utilizing non-volatile RAM as a buffer to improve write latency and offload device management tasks from storage drives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If storage controllers translate data blocks before writing to flash storage, then data management flexibility is improved, but write latency increases and write operations become redundant
Solution Approach 1:
The patent extracts the data translation function from storage controllers and places it directly in the operating system. The OS directly maps host file system addresses to flash storage locations without controller intervention, eliminating redundant translation operations and reducing write latency while maintaining management flexibility through OS-level control
Solution Approach 2:
The patent introduces a buffer memory (DRAM) as an intermediary between the host file system and flash storage. Data is first written to the buffer in the OS, then transferred to flash storage asynchronously, decoupling the write operations and reducing latency while allowing continued data management by the OS
2Adaptability or versatility
If storage controllers manage data allocation and caching, then storage autonomy is improved, but system reliability decreases due to redundant writes
Solution Approach 1:
The patent merges storage management functions (data allocation, caching, translation) from storage controllers into the operating system. This centralization eliminates redundant write operations by having a single authority manage data flow, thereby improving reliability while maintaining storage autonomy through OS-level control mechanisms
Solution Approach 2:
The operating system serves itself by directly managing storage operations without relying on storage controllers. The OS performs its own data translation, buffering, and allocation, eliminating the need for controller-mediated writes that cause redundancy and reliability issues
3Quantity of substance
If multiple storage drives are used to increase capacity, then storage capacity is improved, but write operations become more redundant and performance decreases
Solution Approach 1:
The patent segments the storage system into independent flash storage units that can be directly addressed by the OS. Each unit can be managed independently with direct memory mapping, allowing parallel write operations without the coordination overhead and redundancy that plagues traditional multi-drive controller-managed systems, thereby improving write performance while maintaining capacity
Solution Approach 2:
The patent performs preliminary data preparation in the OS by writing data to buffer memory before final flash storage commitment. This preliminary action allows data to be staged and organized in advance, enabling more efficient multi-drive writes without redundancy when data is finally committed to multiple drives simultaneously
Data Source
AI summary
Data reduction across different cloud-based storage systems, including: performing one or more global data reduction processes on the data using one or more global databases that represent the data associated with the plurality of cloud-based virtual storage systems.