Automated Drive Log Extraction for Fleet Telemetry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage drive logs are designed for failure analysis on individual drives and are ill-suited for telemetry, being too large, producer-unique, and inaccessible with standard commands, lacking time and date information, and not effectively utilized for the vast majority of operational drives.
Innovation Solution
An automated drive log collection and analysis system that synchronizes time and date with power-on hours, adds system-level information to log files, and extracts logs using a common specification, enabling comprehensive analysis of drive workload, performance, and environment across multiple drives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If storage drive logs are designed for failure analysis on individual drives, then detailed drive-specific information is captured, but the logs become too large and inaccessible for telemetry and population-level analysis
Solution Approach 1:
The patent extracts only the necessary telemetry parameters from the full drive logs, creating a separate, simplified log structure designed specifically for population-level analysis. This extraction process separates the telemetry data needed for fleet-wide monitoring from the detailed failure analysis data, reducing overall log size while preserving both types of information in appropriate formats.
Solution Approach 2:
The log structure is segmented into different sections: traditional drive-specific logs for failure analysis and new telemetry parameters for population-level analysis. This segmentation allows both detailed drive information and aggregated telemetry data to coexist without interfering with each other, enabling dual-purpose logging without the complexity of a single monolithic log structure.
2Adaptability or versatility
If drive logs are made producer-unique with nonstandard commands, then drive-specific requirements are met, but standardization and ease of access across multiple drives are compromised
Solution Approach 1:
The patent implements a universal log structure that can accommodate multiple drive types and producers through a standardized interface. The telemetry parameters are defined in a common format that works across different drive manufacturers, while still allowing for drive-specific extensions. This multi-functionality enables both standardized access for population-level analysis and adaptability for drive-specific requirements.
3Reliability
If comprehensive log data is collected from all drives, then complete failure analysis capability is maintained, but the data becomes unwieldy for population-level performance trend analysis
Solution Approach 1:
The patent extracts specific telemetry parameters from comprehensive log data to create a streamlined dataset optimized for population-level analysis. By extracting only the relevant performance metrics, workload characteristics, and environmental parameters needed for fleet-wide trends, the system maintains complete failure analysis capability in the full logs while creating a separate, efficient dataset for productivity-focused population analysis.
Data Source
AI summary
A storage system to communicate with a plurality of storage devices. The storage system includes a processor to execute system software that includes machine readable instructions configured to add system-level information regarding the storage system to log files stored in a reserved area of the storage device, extract the log file from each of the storage devices automatically at a predetermined interval, and transmit the log files from the storage system for analysis.