The design and implementation of a log-structured file system
ACM Transactions on Computer Systems (TOCS)
eNVy: a non-volatile, main memory storage system
ASPLOS VI Proceedings of the sixth international conference on Architectural support for programming languages and operating systems
Using data clustering to improve cleaning performance for plash memory
Software—Practice & Experience
Cleaning policies in mobile computers using flash memory
Journal of Systems and Software
Time Series Analysis, Forecasting and Control
Time Series Analysis, Forecasting and Control
A flash-memory based file system
TCON'95 Proceedings of the USENIX 1995 Technical Conference Proceedings
Storage alternatives for mobile computers
OSDI '94 Proceedings of the 1st USENIX conference on Operating Systems Design and Implementation
Improvement of space utilization in NAND flash memory storages
ICESS'05 Proceedings of the Second international conference on Embedded Software and Systems
LSTAFF: system software for large block flash memory
AsiaSim'04 Proceedings of the Third Asian simulation conference on Systems Modeling and Simulation: theory and applications
Proceedings of the 50th Annual Design Automation Conference
Hi-index | 0.00 |
The problem of flash memory is that it cannot be overwritten unless erased in advance. In order to avoid having to erase during every update, non-in-place-update schemes have been widely used. In case of non-in-place update mechanism, garbage collection is needed to reclaim the obsolete space. In this paper, we study a new garbage collection scheme to reduce its cost such as the number of erase operations and the number of data copies. The proposed scheme determines the victim blocks by exploiting usage information of data blocks such as age, utilization and erase count. In addition, the proposed scheme predicts the future I/O workload and controls the number of victims to avoid disturbing the normal I/O operations. Experimental results show that the proposed scheme can perform well especially when the degree of locality is high.