Multi-path Layer Proactive Path State Updates

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing storage systems lack proactive mechanisms to predict and manage path failures in multi-path input-output (IO) operations, leading to potential IO failures and performance penalties.

Innovation Solution

The implementation of a multi-path layer that identifies health checkpoints for paths between host devices and storage systems, generates health predictions, and proactively updates connection states in multi-path IO drivers when a predicted failure is imminent.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional reactive path failure detection is used, then system complexity remains low, but IO failures and performance penalties occur

Engineering Contradiction:

Improvepath failure prediction accuracyVSAvoidmulti-path layer complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system performs preliminary health predictions and failure date estimations for storage paths before actual failures occur. By proactively identifying paths expected to fail within a threshold period and updating their connection states in advance, the system prevents IO failures and performance penalties rather than reacting after failures happen.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent segments the path health monitoring into discrete checkpoints along each storage path. Each checkpoint can be independently assessed for health status and failure probability, allowing granular control and management of path reliability without requiring complete system-wide complexity.

Inventive Principle:

Principle #1Segmentation

2Productivity

If proactive path state updates are implemented, then IO retries and timeouts are prevented, but processing overhead increases

Engineering Contradiction:

ImproveIO operation throughputVSAvoidhealth prediction and state update time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system implements periodic health predictions at scheduled intervals rather than continuously monitoring every path state change. This periodic approach balances the need for timely failure detection with the constraint of processing overhead, updating path states proactively without constant system intervention.

Inventive Principle:

Principle #19Periodic action

Solution Approach 2:

The multi-path layer automatically performs health predictions and updates connection states without requiring manual intervention or external management system involvement. The system serves itself by autonomously identifying failing paths and adjusting their states, reducing the time and resources needed for manual path management.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS12216526B2Multi-path layer configured for proactive path state changes

Publication Date: 2025.02.04 DELL PROD LP

AI summary

An apparatus comprises a processing device configured to identify, for a given path interconnecting a given host device and a storage system, a set of two or more checkpoints characterizing health of the given path. The processing device is also configured to generate health predictions for each checkpoint in the identified set of two or more checkpoints, and to determine whether any of the generated health predictions indicates an expected date of failure within a designated period of time. The processing device is further configured, responsive to determining that at least one of the generated health predictions indicates an expected date of failure within the designated period of time, to proactively update the connection state of the given path in a given multi-path input-output driver of the given host device.