Offloaded Disaggregated Storage Architecture for SSD Bandwidth Utilization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Disaggregated storage systems fail to effectively utilize the I/O bandwidth of individual solid-state drives (SSDs), leading to suboptimal performance in data management operations.
Innovation Solution
The implementation of a system that offloads data management functions to multiple storage devices within a multi-device storage environment, utilizing a main controller subsystem to determine the appropriate storage devices and operations for data management requests, and performing direct memory access and internal data transfers to enhance data management efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If disaggregated storage systems use traditional architectures (JBOD, AFA, EBOF), then system flexibility and cost savings are achieved, but I/O bandwidth utilization of individual SSDs is insufficient
Solution Approach 1:
The system segments storage operations into two distinct layers: a disaggregated storage layer providing flexible, distributed storage capacity, and an SSD cache layer providing high-speed I/O operations. This segmentation allows each layer to optimize for its specific function while working together to resolve the contradiction between flexibility and performance.
Solution Approach 2:
The patent introduces an intermediary SSD cache layer between the host and the disaggregated storage system. This intermediary captures I/O requests, performs caching operations, and manages data transfers, thereby enhancing I/O bandwidth utilization without compromising the flexibility of the disaggregated storage architecture.
2Device complexity
If data management functions are centralized in the main controller, then system coordination is simplified, but data management throughput is limited
Solution Approach 1:
The patent segments data management functions between the main controller and individual SSDs. The main controller handles high-level coordination and caching decisions, while individual SSDs execute specific data management operations independently, enabling parallel processing and increased throughput.
Solution Approach 2:
The patent implements partial offloading of data management functions to SSDs, where SSDs autonomously handle certain operations (such as cache management and data transfers) without requiring complete controller intervention for each operation, thereby increasing throughput while maintaining manageable system complexity.
3Quantity of substance
If multiple SSDs are used in disaggregated storage, then storage capacity and flexibility increase, but data transfer efficiency decreases
Solution Approach 1:
The patent merges the I/O operations of multiple SSDs into a unified cache layer that presents a single high-performance interface to the host. This merging allows the system to aggregate the capacity of multiple SSDs while maintaining efficient data transfer through coordinated cache operations.
Solution Approach 2:
The patent implements continuous caching operations across multiple SSDs, where the cache layer maintains ongoing data transfers and operations without interruption. This continuity ensures that data transfer efficiency is maintained across the entire disaggregated storage system, leveraging the combined capacity of multiple SSDs.
Data Source
AI summary
Data management functions are offloaded from a main controller to individual storage devices in a multi-device storage environment. The main controller receives a data management request from a host system, and responds by determining one or more storage devices and one or more data management operations to be performed by the one or more storage devices. The main controller initiates performance of a data management function corresponding to the data management request, by sending one or more data management commands to the one or more storage devices, and initiating one or more data transfers, such as a direct memory access operation to transfer data between a memory buffer of a storage device and a host memory buffer of the host system, and an internal data transfer between two or more of the storage devices using an internal communication fabric of the data storage subsystem.