Deadline-Aware I/O Scheduler for QoS Latency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current approaches to ensuring I/O latency meets Quality of Service (QoS) requirements in distributed storage systems are inefficient due to static prioritization methods, which can over-prioritize requests and strain infrastructure unnecessarily.

Innovation Solution

Implementing a deadline-aware I/O scheduler across the cloud provider network components, including compute devices, network devices, and storage devices, that dynamically prioritize I/O requests based on embedded QoS metadata within the I/O request packets, eliminating the need for static priority assignments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If static prioritization of I/O requests is used, then QoS requirements can be met, but infrastructure is over-challenged and efficiency decreases

Engineering Contradiction:

ImproveQoS requirement fulfillmentVSAvoidInfrastructure efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent implements dynamic prioritization where I/O requests are assigned priorities based on real-time deadline urgency rather than static pre-assignment. The scheduler continuously evaluates deadline metadata and adjusts request priorities dynamically, allowing the system to meet QoS requirements only when necessary rather than always challenging infrastructure capacity.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the parameter of request priority from a fixed static value to a dynamic value that varies based on deadline urgency. By calculating urgency factors from deadline metadata and adjusting priorities accordingly, the system optimizes infrastructure utilization while ensuring QoS compliance for time-critical requests.

Inventive Principle:

Principle #35Parameter changes

2Ease of operation

If static prioritization is applied to all requests, then simple scheduling is achieved, but over-prioritization occurs wasting infrastructure capacity

Engineering Contradiction:

ImproveScheduling simplicityVSAvoidInfrastructure capacity waste

Core Design Contradiction:

Ease of operationVSLoss of energy

Solution Approach 1:

The patent applies partial prioritization by assigning high priority only to I/O requests that are at risk of missing their deadlines, rather than uniformly prioritizing all requests. This selective approach maintains scheduling simplicity while avoiding unnecessary infrastructure challenges and capacity waste from over-prioritization.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The system enables I/O requests to self-identify their urgency through embedded deadline metadata, allowing the scheduler to automatically determine which requests require prioritization without complex external control. This maintains operational simplicity while achieving efficient resource allocation.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS11025745B2Technologies for end-to-end quality of service deadline-aware I/O scheduling

Publication Date: 2021.06.01 INTEL CORP

AI summary

Technologies for end-to-end quality of service for I/O operations include a compute device in an I/O path. The compute device receives from another of the compute devices in the I/O path, an I/O request packet. The I/O request packet includes one or more QoS deadline metadata. The QoS deadline metadata is indicative of latency information relating to a currently executing workload relative to a specified QoS. The compute device evaluates the QoS deadline metadata and assigns a priority to the I/O request packet as a function of the evaluated metadata.