Driver Device Mapping Mode Shift for NVM Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory mapped storage implementations face challenges in integrating new non-volatile memory (NVM) technologies due to latency requirements and scalability limitations, necessitating modifications in applications and libraries, as well as block emulation on top of NVM, which complicates data management across different storage types and interconnects.
Innovation Solution
A driver device with a mapping interface and migration unit that dynamically shifts between various mapping modes in real-time, allowing continuous memory mapped access, bypassing software interposed between the application and storage, and enabling data migration between different storage devices and interconnects, thereby simplifying system software environments and hiding differences between storage devices and interconnects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If direct memory mapping of NVM is used to reduce latency, then access speed is improved, but compatibility with legacy systems and scalability are limited
Solution Approach 1:
The system dynamically shifts between legacy and direct mapping modes in real-time based on operational needs. The mapping interface can transition address ranges between different mapping modes without application intervention, allowing the system to optimize for speed when using NVM while maintaining compatibility when accessing traditional storage through the same unified interface.
Solution Approach 2:
The driver device provides a unified memory-mapped interface that handles both legacy block devices and direct NVM access through the same mechanism. This universal interface allows applications to interact with different storage types (traditional disks and NVM) through a single consistent API, eliminating the need for separate code paths while supporting both compatibility and performance modes.
2Adaptability or versatility
If block emulation on top of NVM is implemented to maintain compatibility, then legacy system support is improved, but data management complexity increases
Solution Approach 1:
The patent extracts the mapping mode selection logic from applications and libraries and places it in the driver device. By removing the complexity of managing multiple mapping modes from user-space software and consolidating it in the kernel-level driver, the system maintains legacy compatibility while reducing overall data management complexity for applications.
Solution Approach 2:
The driver device acts as an intermediary between applications and the underlying storage hardware. It provides a unified memory-mapped interface that translates application requests appropriately based on the current mapping mode, whether legacy or direct NVM access, thereby simplifying the interface while managing the complexity of supporting multiple storage types internally.
3Adaptability or versatility
If data migration between storage devices is performed, then storage flexibility is improved, but continuous access without disruption is challenging
Solution Approach 1:
The system performs preliminary actions by allocating new storage resources and setting up the target mapping mode before actually migrating data. The driver device can prepare the destination storage location and configure the mapping interface in advance, then gradually transition data and address ranges while maintaining access through the unified interface, ensuring continuous availability throughout the migration process.
Data Source
AI summary
A plurality of mapping modes may be shifted between in real time while maintaining continuous memory mapped access to an application. Data may be migrated between different types storage devices and/or interconnects. The shift between the plurality of mapping modes may be based on a change to the type storage device and/or type of interconnect for the data migration.