A New Linux Swap System for Flash Memory Storage Devices

Ko, Sohyang; Jun, Seonsoo; Ryu, Yeonseung; Kwon, Oh‐Hoon; Koh, Kern

doi:10.1109/iccsa.2008.54

Cited by 22 publications

(10 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the read speeds from flash are close to those from Nswap. The relative read performance will depend on the particular devices; however, based on this experiment as well as other studies of flash performance [16], [14], [20], we anticipate that reads to flash will rival, and may outperform, reads to Network RAM and that writes to Network RAM will outperform writes to flash. This experiment motivates choosing Network RAM as the initial target of written pages, and then migrating pages from Network RAM to flash so that some subsequent read requests can be satisfied by faster flash; or, in the case when flash and Network RAM are equally good, more reads can be handled in parallel by distributing them over both Network RAM and flash.…”

Section: Motivation For Two Level Driver Systemmentioning

confidence: 79%

Incorporating Network RAM and Flash into Fast Backing Store for Clusters

Newhall

Woos

2011

2011 IEEE International Conference on Cluster Computing

View full text Add to dashboard Cite

Abstract-We present Nswap2L, a fast backing storage system for general purpose clusters. Nswap2L implements a single device interface on top of multiple heterogeneous physical storage devices, particularly targeting fast random access devices such as Network RAM and flash SSDs. A key design feature of Nswap2L is the separation of the interface from the underlying physical storage; data that are read and written to our "device" are managed by our underlying system and may be stored in local RAM, remote RAM, flash, local disk or any other cluster-wide storage. Nswap2L chooses which physical device will store data based on cluster resource usage and the characteristics of various storage media. In addition, it migrates data from one physical device to another in response to changes in capacity and to take advantage of the strengths of different types of physical media, such as fast writes over the network and fast reads from flash. Performance results of our prototype implementation of Nswap2L added as a swap device on a 12 node Linux cluster show speed-ups of over 30 times versus swapping to disk and over 1.7 times versus swapping to flash. In addition, we show that for parallel benchmarks, Nswap2L using Network RAM and a flash device that is slower than Network RAM can perform better than Network RAM alone.

show abstract

Section: Motivation For Two Level Driver Systemmentioning

confidence: 79%

Incorporating Network RAM and Flash into Fast Backing Store for Clusters

Newhall

Woos

2011

2011 IEEE International Conference on Cluster Computing

View full text Add to dashboard Cite

show abstract

“…It adopts an existing garbage collection policy called cost-benefit algorithm, which does not consider the degree of wear leveling during the selection of victim blocks for erasure. Ko et al proposed a new Linux swap system called LOBI [4], which adopts a block-aligned swap-in read-ahead algorithm to reduce the number of cleaned blocks and delay the execution of garbage collection operation. Lin et al proposed a Linux swap system, called SGBI [1], where Linux kernel 2.6 manages the flash memory-based swap space directly without FTL.…”

Section: A Existing Work On Swap Space Management Schemementioning

confidence: 99%

“…Due to limited memory resource of portable consumer electronics, portable consumer electronics currently exploit efficient Linux swap systems [1]- [4] considering flash memory as swap space as a cost effective solution to extend limited memory space. The flash memory-based swap space is always used up in a short period of time due to the out-of-place update scheme used to solve the erase-before-write constraint of flash memory.…”

Section: Introductionmentioning

confidence: 99%

Flash-aware linux swap system for portable consumer electronics

Lin

Chen

2012

IEEE Trans. Consumer Electron.

View full text Add to dashboard Cite

Because of limited memory resource of portable consumer electronics, portable consumer electronics currently exploit swap space with flash memory as a cost effective solution to extend limited memory space. Original swap space management scheme used in Linux for magnetic disk is not available to flash memory-based swap space due to the distinct characteristics of flash memory.In this paper, we propose a flash-aware Linux swap system, called FLSS, which adopts Linux kernel 2.6 to manage flash memory-based swap space directly without FTL. We introduce: 1) a partial block alignment scheme to perform an efficient swap-in read-ahead algorithm, 2) a swap-aware victim block selection method and the redefined concept of hot page and cold page to design a swap-aware garbage collection policy called SACATA, and 3) the notion of overage as well as the notion of frozen applied to blocks to exploit a wear leveling-aware block management scheme. Experimental results show that the proposed FLSS greatly outperforms existing swap space management techniques and evaluate the effectiveness of proposed SACATA 1 .

show abstract

“…To achieve these design principles, Ko et al proposed a new Linux swap system called LOBI [1], which eliminates the cleaning cost of garbage collection in terms of the reduced number of cleaned blocks by adopting a block-aligned swap-in read-ahead scheme. But it does not propose an efficient garbage collection policy to eliminate the cleaning cost of single garbage collection and improve the degree of wearlevelling.…”

mentioning

confidence: 98%