Storage Area Expiration Grouping for Faster Garbage Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage systems face inefficiencies in garbage collection due to randomly distributed expiration times of recovered data, leading to prolonged migration times and reduced efficiency.
Innovation Solution
The method involves storing valid data in storage areas with centralized expiration times, ensuring that all data in a storage area either becomes invalid or remains valid simultaneously, thereby reducing the need for data migration during garbage collection and improving collection efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If recovered data is randomly stored in storage areas, then data can be stored flexibly, but garbage collection time increases and efficiency decreases
Solution Approach 1:
The patent segments storage areas into multiple units and assigns recovered data to specific storage areas based on expiration time characteristics. This segmentation allows the system to group data with similar expiration times together, enabling more efficient garbage collection by processing entire storage areas as uniform units rather than individually managing scattered data across the storage system.
Solution Approach 2:
The patent changes the storage parameter from random distribution to expiration time-based distribution. By using expiration time as the key parameter for determining storage area assignment, the system transforms the storage organization from a flexible but inefficient random approach to a structured approach where data with similar expiration characteristics are grouped together, significantly reducing garbage collection time.
2Quantity of substance
If valid data with different expiration times is stored in the same storage area, then storage space utilization is improved, but data migration during garbage collection increases
Solution Approach 1:
The patent applies local quality by creating storage areas with uniform expiration time characteristics. Instead of having heterogeneous data with varying expiration times in each storage area, the system ensures that each storage area contains data with similar expiration properties. This local uniformity reduces the need for data migration during garbage collection, as entire storage areas can be processed as uniform units.
Solution Approach 2:
The patent performs preliminary action by organizing data into storage areas with uniform expiration times before garbage collection occurs. This pre-organization based on expiration time characteristics eliminates the need for complex data migration during garbage collection, as the system can directly identify and process entire storage areas that are ready for collection without moving individual data blocks.
3Ease of operation
If data is migrated during garbage collection, then storage areas can be reused, but the process becomes time-consuming and inefficient
Solution Approach 1:
The patent performs preliminary organization of data into storage areas with uniform expiration times before garbage collection is needed. This pre-organization allows the system to quickly identify storage areas ready for reuse without time-consuming data migration during the garbage collection process itself, as entire storage areas can be reclaimed as uniform units.
Solution Approach 2:
The patent uses copying to create storage area metadata that tracks expiration time characteristics. By maintaining copies of expiration time information in the storage area metadata, the system can quickly identify and process storage areas for garbage collection without examining individual data blocks, significantly reducing the time required for storage area reuse.
Data Source
Figure 1
Figure 2
Figure 3~5
AI summary
A data storage method and apparatus, and a storage system are provided. The method is applied to a storage system. The storage system includes at least one first memory. The storage system further includes a second memory. The at least one first memory includes a plurality of storage areas. Garbage collection is performed based on each of the plurality of storage areas. In the method, a time length between the earliest expiration time and the latest expiration time in expiration time of valid data stored in a target storage area is limited to be less than or equal to a preset time length, so that the expiration time of the valid data stored in the target storage area is comparatively centralized. This helps reduce time occupied for migrating the valid data, and improve efficiency of the garbage collection in a process of performing the garbage collection in the target storage area.