Partial Map Exception Tracking for Shingled Write Failures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage devices face challenges in accurately positioning heads over disk tracks due to errors in reading servo bursts, leading to incomplete writes and potential data loss during power failures or read-modify-write operations.
Innovation Solution
The implementation of a control circuitry that generates a partial map to identify and update exception entries for partially written data sectors, allowing for retry operations and dynamic LBA mapping to ensure data integrity and validity, even in the presence of power failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If read-modify-write operations are performed during shingled magnetic recording, then data integrity can be maintained through exception tracking, but device complexity increases due to the need for partial map generation and exception entry management
Solution Approach 1:
The patent segments the data sector into multiple host blocks (e.g., four 512-byte blocks per sector) and implements fine-grained exception tracking at the host block level rather than sector level. The exception entry includes a bit map where each bit corresponds to a specific host block, allowing selective marking of invalid blocks. This segmentation enables partial recovery scenarios where some blocks remain valid even if others fail during read-modify-write operations.
Solution Approach 2:
The patent introduces an intermediary data structure called a 'partial map' that serves as a temporary storage mechanism between the read-modify-write operation and the exception entry update. The partial map holds the location information of newly written host blocks and is used to update the exception entry's bit map. This intermediary structure simplifies the complexity by providing a clear separation between the operation execution and the exception state management.
2Measurement precision
If exception entries track partially written sectors with bit maps, then data validity can be determined for individual host blocks, but measurement precision requirements increase for tracking block-level exceptions
Solution Approach 1:
The patent applies local quality by creating heterogeneous data structures with different levels of detail for different purposes. The exception entry contains a bit map for fine-grained host block tracking (local level) and references to partial maps for sector-level operations (global level). This allows the system to use the appropriate level of precision depending on the specific operation, reducing unnecessary complexity while maintaining the required measurement precision for block-level exception tracking.
3Reliability
If the read operation fails during read-modify-write, then data loss can be prevented by generating exception entries, but productivity decreases due to operation failures and retry requirements
Solution Approach 1:
The patent implements preliminary action by pre-allocating exception entry structures and partial map buffers before read-modify-write operations begin. The control circuitry is pre-configured with the exception entry template including the bit map structure, and partial map storage is pre-allocated in memory. This preliminary preparation eliminates the need for complex runtime allocation and structure initialization during failure recovery, reducing the productivity penalty associated with error handling.
Solution Approach 2:
The exception entry system operates autonomously through self-service mechanisms. When a read-modify-write operation fails, the control circuitry automatically generates the exception entry, updates the bit map based on the partial map data, and marks affected host blocks as invalid without requiring external intervention. The system also automatically handles retry logic by checking exception entries before subsequent operations, allowing failed operations to be recovered without manual intervention and maintaining productivity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This solution ensures that data storage devices can recover from incomplete writes and maintain data integrity by tracking valid and invalid host blocks, enabling efficient access and error handling during read-modify-write operations and power failures.
Implementation Method 1
a head actuated over a disk, wherein the head is connected to a distal end of an actuator arm which is rotated about a pivot by a voice coil motor (VCM)
Implementation Method 2
The phase based servo bursts 14 provide fine head position information used for centerline tracking while accessing a data track during write/read operations. A position error signal (PES) is generated by reading the servo bursts 14
Data Source
AI summary
A data storage device is disclosed wherein a first host block is written to a first data sector, and when writing a second host block to a second data sector the first host block is read from the first data sector. When the read of the first host block fails, a partial map is generated identifying a location of the second host block in the second data sector, the partial map is stored in a non-volatile memory, and the second host block is written to the second data sector. When a power failure occurs after writing the second host block to the second data sector, an exception entry is updated using the partial map, wherein the exception entry is associated with the first and second host blocks.