Object ID Versioning for Reducing Metadata Volume in Object-Based Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In object-based storage systems, frequent updates to files lead to a large number of object IDs being generated, occupying significant storage space and increasing system resource usage, as each update requires a new object ID even if the data's offset range remains unchanged.
Innovation Solution
The system uses a combination of an object ID and a version number to manage updates, where the object ID remains unchanged if the offset range of the updated data does not change, reducing the total number of object IDs and minimizing metadata synchronization needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a new object ID is generated for each file update, then data versioning is tracked accurately, but the total quantity of object IDs becomes very large, occupying relatively large storage space and increasing system resource loss
Solution Approach 1:
The patent segments the object identification into two parts: object ID and version number. The object ID remains stable for a given object, while the version number tracks updates. This segmentation allows the system to maintain accurate versioning information without generating a new unique object ID for each update, thereby reducing the total quantity of object IDs while preserving data versioning tracking capability
Solution Approach 2:
The patent adds a version number dimension to the object identification system. Instead of relying solely on object IDs to track versions, the system introduces a separate version number parameter that evolves independently. This dimensional addition allows the system to maintain precise version tracking while keeping object IDs stable, thus resolving the contradiction between versioning reliability and object ID quantity
2Loss of information
If a new object ID is generated for each file update, then data modification is recorded completely, but metadata synchronization complexity and system resource usage increase
Solution Approach 1:
By segmenting the identification system into object ID and version number, the patent reduces metadata synchronization complexity. When an object is updated, only the version number needs to be incremented and synchronized, rather than synchronizing a new object ID. This segmentation maintains complete modification recording while reducing the complexity of metadata synchronization
Solution Approach 2:
The patent extracts the version tracking function from the object ID and places it in a separate version number parameter. This extraction allows the object ID to remain stable and simplifies synchronization, as only the version number needs to be updated and propagated across the system to record data modifications, thereby reducing metadata synchronization complexity
Data Source
AI summary
Embodiments of the present invention provide a data management technology. An OSD receives a strip write request sent by a client server, where the strip write request carries a to-be-written strip, a version number of the to-be-written strip, an offset of the to-be-written strip, and an object ID of the to-be-written strip; and the OSD writes the to-be-written strip into a storage location determined by using the object ID, the version number of the to-be-written strip, and the offset of the to-be-written strip. By applying the present invention, a quantity of object IDs can be reduced.