SDS Direct I/O Execution for Multi-Hop Disaggregated Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Disaggregated infrastructure systems face performance and resource utilization issues due to the need for I/O commands to move through multiple 'hops' requiring processing resources, memory resources, and network bandwidth when data is forwarded and copied.
Innovation Solution
A Software-Defined Storage (SDS)-enabled system that translates I/O commands with direct memory access information, allowing direct execution between a computing system and a storage system, eliminating the need for intermediate processing and network bandwidth usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If storage resources are disaggregated from LCSs, then flexibility and independent scaling are improved, but I/O command processing complexity and resource consumption increase due to multiple hops
Solution Approach 1:
The SDS engine acts as an intermediary component that receives I/O commands from the LCS, translates them into direct I/O commands, and enables direct execution without requiring data to pass through multiple intermediate hops. This mediator approach maintains the flexibility of disaggregated storage while simplifying the I/O command processing path.
2Adaptability or versatility
If data is forwarded through multiple hops in disaggregated infrastructure, then storage resource flexibility is improved, but network bandwidth and processing resources are consumed
Solution Approach 1:
The invention extracts the data transfer path from the traditional multi-hop routing by enabling direct I/O commands to access storage resources without forwarding data through intermediate network hops. This eliminates unnecessary network bandwidth consumption while preserving the flexibility to access disaggregated storage resources.
Solution Approach 2:
The direct I/O execution mechanism allows the storage system to serve itself by directly executing I/O commands from the LCS without requiring intermediate processing and data forwarding through multiple hops, thereby reducing network bandwidth consumption while maintaining storage flexibility.
3Adaptability or versatility
If I/O commands pass through multiple hops, then storage disaggregation flexibility is maintained, but processing resources and memory resources are required at each hop
Solution Approach 1:
The SDS engine serves as a mediator that consolidates the I/O command translation and execution function, eliminating the need for multiple intermediate processing hops. This reduces the processing resources and memory resources required across the system while maintaining the flexibility of storage disaggregation.
Data Source
AI summary
A Software-Defined Storage (SDS)-enabled disaggregated infrastructure direct Input/Output (I/O) execution system includes an SDS subsystem that is coupled to each of a computing system and a storage system. The SDS subsystem receives an I/O command from the computing system that is directed to the storage system and that includes computing system direct memory access information. The SDS subsystem translates the I/O command to provide a translated I/O command. The SDS subsystem then provides the translated I/O command along with the computing system direct memory access information to the storage system. The storage system may then execute the translated I/O command directly with the computing system using the computing system direct memory access information.