Copy-Free Filesystem Migration via Metadata Transfer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data migration across filesystems with different formats is disruptive, time-consuming, and requires copying all user data, leading to resistance from storage network administrators due to suboptimal utilization and inability to leverage new features in upgraded releases.
Innovation Solution
The technology facilitates copy-free data migration by using node computing devices to transfer filesystem metadata, allowing seamless transition of data ownership from a source filesystem to a destination filesystem without copying user data, utilizing a migration management module to manage the process and ensure compatibility across different addressing schemes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data is copied during filesystem migration, then data can be transferred between different filesystem formats, but migration time and disruption increase significantly
Solution Approach 1:
The patent uses metadata copying instead of data copying. Filesystem metadata including pointers and references to user data blocks are copied from source to destination filesystem, allowing the migration to proceed by updating references rather than physically moving data, thus dramatically reducing migration time while maintaining compatibility between different filesystem formats
Solution Approach 2:
The patent introduces a migration management module as an intermediary that coordinates the migration process. This module manages the copying of metadata, updates pointers and references, and ensures data integrity during the transition between filesystem formats, enabling seamless migration without direct data copying
2Adaptability or versatility
If all user data is copied during migration, then data can be migrated across filesystems with different formats, but additional storage space is required
Solution Approach 1:
The patent applies copying selectively to metadata structures rather than user data. By copying only filesystem metadata, pointers, and references during migration, the system achieves format conversion between different filesystems without requiring additional storage space for duplicating actual user data, thus resolving the space constraint while maintaining format adaptability
3Adaptability or versatility
If data migration is performed using traditional copying methods, then data can be transferred between filesystems, but application disruptions and time-outs occur
Solution Approach 1:
The patent uses metadata copying that allows applications to continue accessing data during migration. By updating pointers and references in the metadata rather than moving data blocks, the system maintains data accessibility throughout the migration process, ensuring application continuity and eliminating disruptions while still enabling data transfer between filesystems
Solution Approach 2:
The migration management module performs preliminary setup by establishing the destination filesystem structure and copying metadata before actual migration operations. This preliminary action includes preparing pointer updates and reference mappings, which enables seamless data access during the transition and prevents application time-outs by ensuring data remains accessible throughout the process
Data Source
AI summary
Methods, non-transitory computer readable media, and computing devices that facilitate copy-free data migrations across filesystems. In a first step with this technology, a first set of filesystem metadata associated with a first filesystem is received. At least a portion of the first set of filesystem metadata is retrieved from a first data structure associated with the first filesystem. The first set of filesystem metadata includes a first identifier and a physical location associated with user data. A second identifier, associated with a second filesystem having a different addressing scheme than the first filesystem, is generated from at least the first identifier. A second set of filesystem metadata including the second identifier and the physical location is stored such that at least the second identifier is stored in a second data structure associated with the second filesystem.