Storage-Hosted Application Acceleration for Network Latency Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The slow execution of important software due to latency caused by large numbers of storage access requests across data-center networks, which is a significant challenge in handling exa-scale datasets and network latency in current datacenter architectures.
Innovation Solution
The system enables application software to be hosted and executed on storage devices equipped with Customizable Storage Controllers (CSC), reducing the need for network round trips by running storage-facing portions of the application software directly on the storage devices, thereby minimizing network latency and congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If application software is executed on servers with network-connected storage devices, then system architecture is simple and ease of operation is maintained, but network latency and congestion increase significantly reducing execution speed
Solution Approach 1:
The patent merges the application software execution environment directly with the storage device by hosting the software on the storage device's controller. This combination eliminates the need for separate server-compute nodes to communicate with storage over the network, thereby reducing network latency and accelerating data access operations.
Solution Approach 2:
The storage device controller acts as an intermediary that hosts and executes application software locally. Instead of servers remotely accessing storage, the controller mediates by running the software that directly interacts with the stored data, eliminating network round-trips and reducing congestion.
2Productivity
If application software is hosted on storage devices, then network traffic and latency are reduced improving speed, but device complexity increases
Solution Approach 1:
The storage device controller is designed with multi-functionality, serving both as a storage management interface and as a computing platform capable of hosting and executing application software. This universal design allows the same hardware to perform multiple roles without requiring separate dedicated systems.
Solution Approach 2:
The storage device architecture is made dynamic by enabling flexible software hosting capabilities. The controller can dynamically load, execute, and manage application software alongside its storage functions, adapting to different workloads and data access patterns without fixed hardware configurations.
3Speed
If more network bandwidth is allocated for storage access, then data transfer speed improves, but network congestion increases and latency worsens
Solution Approach 1:
The patent extracts data access operations from the network layer by hosting application software directly on the storage device. This extraction removes the need for continuous network communication for data access, thereby eliminating network congestion and performance instability caused by high storage traffic.
Data Source
AI summary
A storage device management system (SDMS) as a component of an application acceleration system; this is for managing AppAccel hardware storage devices (AI hardware SDs) using a distributed hardware F+ Update Server service layer of one or more hardware F+ Update Servers managing a storage set of one or more AI hardware SDs. The one or more hardware F+ Update Servers manage the one or more AI hardware SDs with mechanisms such as SyncLink, data paths, SC Program Store, etc. having multiple uses among which is interoperating with AI hardware SD management components of the one or more hardware F+ Update Servers. The SDMS is composed of the one or more hardware F+ Update Servers and employ these mechanisms on one, many, or all of the one or more AI hardware SDs to enable feedback to the SDMS, which is an important operative part of managing one, many, or all of the one or more AI hardware SDs.