Global File Namespace Mapping for Multi-File-System Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The capacity expansion and management of file systems in data centers are complex due to differences between vendors and models, leading to cumbersome data migration processes that require significant manpower and resources.
Innovation Solution
A global name space is constructed to map all file paths in multiple file systems to a unique path, decoupling the service side from the file systems, allowing transparent management and flexible configuration, with the use of a mount information table to determine file paths and storage addresses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data is migrated from old file system to new file system during capacity expansion, then the service can use the new file system, but the migration process consumes a large amount of manpower and material resources
Solution Approach 1:
The patent introduces a name space as an intermediary layer between the service and multiple file systems. This name space maps file paths from different file systems (old and new) to a unified structure, allowing the service to access data without direct migration. The mount information table acts as a mediator that translates service requests to the appropriate file system locations, eliminating the need for complex data migration while maintaining compatibility.
2Ease of operation
If a global name space is constructed to map file paths across multiple file systems, then service modifications are eliminated during file system changes, but the name space management complexity increases
Solution Approach 1:
The name space is designed with multi-functionality to handle diverse file system types and path structures through a unified interface. It can map paths from different vendors and models simultaneously, providing universal access without requiring service modifications. The mount information table stores correspondence relationships that enable this universal mapping capability.
3Quantity of substance
If different file systems from different vendors and models are used in data center capacity expansion, then storage capacity increases, but the file systems become independent and data migration becomes necessary
Solution Approach 1:
The patent segments the file system access layer from the storage layer by introducing a name space. This allows different file systems (from different vendors and models) to remain independent at the storage layer while being integrated at the access layer through path mapping. Each file system can be independently managed and expanded, yet all are accessible through the unified name space structure.
Data Source
AI summary
A method includes: obtaining an access request, wherein the access request carries a first file path corresponding to to-be-accessed data, and the first file path comprises a preset root directory; determining, based on a path of the first file path in the preset root directory, a second file path in a first file system corresponding to the first file path, wherein a name space corresponding to the preset root directory comprises at least two types of sub-paths, and the at least two types of sub-paths respectively correspond to file paths in at least two different file systems; and determining a storage address of the to-be-accessed data based on the second file path.