HSM Data Migration Preserving Layout via Stub File Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data migration methods using standard file system calls fail to distinguish between stub files and ordinary files, leading to incorrect data layout reproduction and increased storage costs when migrating data between HSM execution environments, as they alter time attributes and require unnecessary storage allocation.
Innovation Solution
Implementing a data migration function that checks if files are stub files or ordinary files using extended file system functions, converts stub files, and reads data from secondary storage to maintain the original data layout by reserving equivalent primary storage space in the destination, allowing data to be migrated while preserving the data layout structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If standard file system calls are used for data migration, then data can be migrated between storage devices, but the data layout structure is altered and storage costs increase due to incorrect reproduction of hierarchical storage management
Solution Approach 1:
The patent introduces a migration source file server and migration destination file server as intermediaries that understand HSM-specific file structures. These servers act as mediators between the source and destination storage devices, enabling data migration while preserving the hierarchical layout by interpreting and reconstructing stub file relationships at the file server level rather than through generic file system calls.
Solution Approach 2:
The patent changes the approach from generic file system parameters to HSM-specific parameters by introducing flags to identify stub files versus regular files. This parameter change enables the migration system to distinguish between file types and apply appropriate migration handling, thereby preserving the data layout structure through awareness of the hierarchical storage management semantics.
2Ease of operation
If all data is allocated in primary storage during migration, then data can be migrated to HSM environment, but storage costs increase due to unnecessary storage allocation
Solution Approach 1:
The patent segments the storage allocation strategy by treating stub files and regular files differently. Instead of allocating all data in primary storage, the system segments the migration process to allocate only stub files in primary storage and regular files in secondary storage, matching the destination HSM environment's hierarchical structure and avoiding unnecessary primary storage consumption.
Solution Approach 2:
The patent performs preliminary identification of file types (stub files vs. regular files) during the migration planning phase. This preliminary action enables the system to pre-determine the appropriate storage allocation for each file type, ensuring that files are directed to the correct storage tier before actual data transfer, thereby optimizing storage space utilization.
3Productivity
If time attributes are altered during standard file migration, then data can be transferred, but the original data access environment is changed affecting data retrieval behavior
Solution Approach 1:
The patent uses a copying mechanism that preserves original file attributes. Instead of transferring files through operations that modify timestamps, the system copies files while explicitly maintaining their original time attributes, ensuring that the data access environment remains unchanged and retrieval behavior is preserved.
Solution Approach 2:
The migration system implements self-service by automatically preserving and restoring file attributes during the migration process. The file servers handle attribute preservation without external intervention, ensuring that time attributes and other metadata remain intact throughout the migration, thereby maintaining data access environment stability.
Data Source
AI summary
When a migration object is a stub file, a data migration apparatus, which migrates data from a source HSM execution environment to a destination HSM execution environment, reads data from source secondary storage based on a data storage address in the source secondary storage. This data storage address is stored in the stub file in source primary storage. The apparatus writes the data into destination secondary storage according to a rule for storing data and reads attribute information on the stub file stored in the source primary storage. The apparatus creates, in destination primary storage, a stub file having the same name as in the source and, according to the stub file generation format in the destination primary storage, supplies the address of the data stored in the destination secondary storage and attribute information acquired from the source primary storage to the created stub file to complete the data migration.