Host Accelerator Deduplication Signatures for Storage Load Reduction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Storage systems experience performance degradation due to the substantial computational resources consumed in computing hashes or content-based signatures for data written by host devices, leading to increased load.

Innovation Solution

Leverage host accelerator resources, such as GPUs and DPUs, to compute content-based signatures for data written to storage systems, reducing the computational burden on storage arrays.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the storage system computes content-based signatures for data blocks, then data deduplication functionality is enabled, but computational resources and processing load increase significantly

Engineering Contradiction:

Improvedata deduplication functionalityVSAvoidcomputational resources

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts the computationally intensive signature computation task from the storage system and relocates it to host devices. The storage system only receives pre-computed signatures from hosts, thereby eliminating the computational burden while preserving deduplication functionality. This is achieved through host-side agents that compute signatures before data transfer.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces host devices as intermediaries between the data source and storage system. These intermediaries perform the signature computation on behalf of the storage system, acting as a mediator that relieves the storage system of computational tasks while enabling deduplication through the transmitted signatures.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If the storage system processes multiple host devices with varying workloads, then service coverage is improved, but performance degradation occurs due to computational overhead

Engineering Contradiction:

Improveservice coverageVSAvoidstorage system performance

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent implements self-service by enabling host devices to autonomously compute their own data signatures using local acceleration resources. Each host device independently performs signature computation without burdening the storage system, allowing the storage system to serve multiple hosts simultaneously without performance degradation from computational overhead.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS20250298507A1Leveraging host accelerator resources to reduce storage system computational load

Publication Date: 2025.09.25 DELL PROD LP

AI summary

An apparatus illustratively comprises at least one processing device that includes a processor and a memory, with the processing device being configured to obtain in a host device deduplication information relating to a logical storage device of a storage system, wherein the host device comprises acceleration resources and is configured to communicate with the storage system over at least one network, and to determine based at least in part on the obtained deduplication information whether to compute a content-based signature for one or more data blocks of the logical storage device utilizing the acceleration resources of the host device. Responsive to an affirmative result of the determining, the content-based signature is computed utilizing the acceleration resources of the host device and sent by the host device to the storage system in association with a given write operation that targets the one or more data blocks of the logical storage device.