Drive Retry Time Tracking for Storage System Stability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current disk drive failure prediction methods in data storage systems are inadequate as they rely solely on counter-based retry thresholds without considering response time, leading to potential system panics and require extensive error handling knowledge, failing to proactively manage drives that take excessively long to respond.
Innovation Solution
The implementation of a mechanism where the physical drive object tracks overall retry time and initiates health checks, allowing the RAID object to proactively handle drive failures by considering time-based retry policies and simplifying the interface for error handling, enabling the system to determine whether an I/O can be retried and how long to wait before retrying, and performing self-tests to recover or replace faulty drives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If counter-based retry thresholds are used to manage drive failures, then error monitoring is simplified, but response time is not considered leading to potential system panics
Solution Approach 1:
The patent changes the parameter from counter-based retry thresholds to time-based retry thresholds. Instead of counting the number of retry attempts, the system now monitors the total time spent on retry operations. This parameter change allows the system to consider both error counters and response times, preventing system panics by timing out slow drives while maintaining operational simplicity.
2Reliability
If extensive error handling knowledge is required to manage drive failures, then precise error management is achieved, but device complexity increases
Solution Approach 1:
The patent implements self-service by allowing the drive itself to report its status through the physical drive object, which automatically determines whether an I/O can be retried and tracks retry time. The system automatically takes offline drives that exceed the retry time threshold without requiring extensive external error handling knowledge or complex management procedures.
3Reliability
If drives are taken offline immediately upon error detection, then data corruption is prevented, but downtime increases
Solution Approach 1:
The patent applies preliminary action by tracking the overall retry time before taking offline action. Instead of immediately taking a drive offline upon error detection, the system allows for a predetermined retry period. If the drive does not respond within this time threshold, it is then taken offline. This prevents premature offline actions while still protecting against data corruption from unresponsive drives.
Data Source
AI summary
A method is used in managing drive problems in data storage systems. It is determined, by a physical drive object, whether an I/O with a drive is able to be retried by another object. An overall retry time for the I/O is tracked. If the overall retry time is excessive, the drive is taken offline.