Host Virtual Space Mapping for Distributed Storage Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing file access methods for non-volatile memory-based storage systems, such as those using NFS or CIFS protocols, result in low access efficiency due to the need for kernel TCP/IP protocol stack processing, which increases delays and reduces performance.
Innovation Solution
A file access method that involves a host obtaining information about a target file's virtual space in a distributed storage system, allocating a local virtual space, establishing a correspondence between the virtual and global addresses, and converting access requests to directly route data to the appropriate storage node, thereby bypassing the kernel and improving access efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing file access methods (NFS/CIFS) are used for NVM-based storage systems, then compatibility with conventional file systems is maintained, but access efficiency deteriorates due to kernel TCP/IP protocol stack processing
Solution Approach 1:
The patent extracts the file access operation from the conventional kernel TCP/IP protocol stack and relocates it to the storage node's kernel space. The host initiates access requests that are intercepted and processed directly by the storage node's kernel, bypassing the network protocol stack processing that occurs in the host kernel. This extraction of the access path removes the bottleneck caused by kernel-space protocol processing while maintaining the distributed file system architecture.
Solution Approach 2:
The patent introduces a virtual file system (VFS) layer as an intermediary between the host's user space applications and the storage node's kernel space. The VFS layer on the storage node receives access requests from the host, translates them into appropriate storage operations, and directly accesses the NVM devices. This intermediary layer enables efficient direct access while maintaining compatibility with standard file system interfaces.
2Productivity
If direct access to physical storage is implemented, then access efficiency is improved, but system complexity increases due to address space management
Solution Approach 1:
The patent introduces a virtual address space dimension to manage the complexity of direct physical storage access. Instead of managing physical addresses directly, the system creates a virtual address space layer that maps to physical storage locations. The VFS layer maintains virtual file descriptors and address mappings, transforming complex physical address management into simpler virtual address management that is easier to handle and more flexible.
Solution Approach 2:
The patent creates virtual copies of file access interfaces in the VFS layer that mirror the semantics of conventional file system operations. These virtual file descriptors and access interfaces replicate the behavior of standard file operations, allowing applications to interact with NVM-based storage through familiar interfaces while the underlying physical access is optimized. This copying approach hides the complexity of direct physical access behind simplified virtual interfaces.
Data Source
AI summary
A file access method, a system, and a host are provided. According to the method, after obtaining information about first virtual space of a target file, a host allocates, in local virtual address space of the host, second virtual space to the target file, where the first virtual space is space allocated in global virtual address space by a management node in a distributed storage system to the target file. The host converts, according to a correspondence between the first virtual space and the second virtual space, a second access request of accessing the second virtual space into a first access request, where an address of the first virtual space in the first access request includes device information of a first storage node. Then, the host sends the first access request to a network device to route the first access request to the first storage node.