Magnetic Disk Trace Data Storage for Failure Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic disk devices often fail during use, making it difficult to investigate the cause of failure due to changes in signal quality when using an expensive bus analyzer, and sometimes access is restricted, limiting inspection capabilities, while SMART information is insufficient for identifying interface-related issues.
Innovation Solution
A magnetic disk device configuration with a system-on-chip (SoC) that executes read/write processing and stores trace data, allowing for detailed analysis of failures without an expensive bus analyzer, enabling investigation of failure causes at low cost and without affecting I/O throughput.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a bus analyzer is attached to investigate failure causes, then measurement precision is improved, but device complexity increases and signal quality changes causing inability to reproduce errors
Solution Approach 1:
The patent creates a copy of the interface communication data by having the magnetic disk device itself store trace data of read/write processing. This eliminates the need to attach external bus analyzers, avoiding signal quality changes while maintaining measurement capability. The trace data serves as a copy of the actual interface communications that can be analyzed later without affecting the original signal environment.
Solution Approach 2:
The magnetic disk device performs self-diagnosis by storing its own interface communication trace data internally. Instead of requiring external analysis equipment, the device serves its own investigation needs by capturing and storing the actual communication data between the host and the device, enabling failure analysis without external interference.
2Measurement precision
If a bus analyzer is attached to investigate failure, then measurement precision is improved, but loss of time increases due to inability to reproduce errors
Solution Approach 1:
The trace data is stored in advance during normal operation, capturing the actual interface communications before failures occur. This preliminary recording eliminates the need for repeated real-time monitoring and error reproduction attempts, significantly reducing investigation time while maintaining measurement accuracy.
3Loss of information
If SMART information is read to investigate failure, then loss of cost is reduced, but measurement precision deteriorates due to limited information content
Solution Approach 1:
The patent segments the information storage into two parts: traditional SMART data for basic health monitoring and detailed trace data for specific interface communication analysis. This segmentation allows investigators to use the detailed trace data for precise interface-related failure identification while maintaining the cost benefits of internal device storage.
4Measurement precision
If trace data is stored during read/write processing, then measurement precision is improved for failure analysis, but productivity decreases due to potential interruption of data transfer
Solution Approach 1:
The trace data storage operates continuously alongside normal read/write processing without interrupting data transfer. The controller simultaneously manages both the primary data I/O operations and the trace data recording, ensuring that investigation capabilities are maintained while productivity remains uninterrupted.
Data Source
AI summary
According to one embodiment, a magnetic disk device incudes a first magnetic disk, a first actuator that reads/writes data from/to the first magnetic disk, a first controller that controls the first actuator, a second magnetic disk, a second actuator that reads/writes data from/to the second magnetic disk, a second controller that controls the second actuator, wherein the first controller executes read/write processing, and the second controller stores trace data of the read/write processing when the first controller executes the read/write processing.