BW-DCache: an inexpensive, effective and reliable cache solution in a SAN file system

Authors:
Chengxiang Si;Xiaoxuan Meng;Lu Xu
Affiliations:
Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China;Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China;Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China
Venue:
HPCS'09 Proceedings of the 23rd international conference on High Performance Computing Systems and Applications
Year:
2009

Citing 10
Cited 0

Scale and performance in a distributed file system

ACM Transactions on Computer Systems (TOCS)
Modern operating systems

Modern operating systems
Cache Memories

ACM Computing Surveys (CSUR)
Properties of the working-set model

Communications of the ACM
Hint-based cooperative caching

ACM Transactions on Computer Systems (TOCS)
2Q: A Low Overhead High Performance Buffer Management Replacement Algorithm

VLDB '94 Proceedings of the 20th International Conference on Very Large Data Bases
The Multi-Queue Replacement Algorithm for Second Level Buffer Caches

Proceedings of the General Track: 2002 USENIX Annual Technical Conference
STICS: SCSI-to-IP cache for storage area networks

Journal of Parallel and Distributed Computing
ARC: A Self-Tuning, Low Overhead Replacement Cache

FAST '03 Proceedings of the 2nd USENIX Conference on File and Storage Technologies
Cooperative caching: using remote client memory to improve file system performance

OSDI '94 Proceedings of the 1st USENIX conference on Operating Systems Design and Implementation

Quantified Score

Hi-index	0.01

Visualization

Abstract

In order to enhance scalability of Blue Whale Cluster File System (BWFS), this paper analyses the bottleneck of scalability of BWFS in non-linear editing application, and presents a solution that clients employ local disk storage as cache on clients. We called BW-DCache (BWFS Cache based Disk medium). Moreover, we carefully choose the client cache strategy and replacement algorithm according to data access pattern. By this, clients can get as much data from local cache as possible, avoid random accesses to storage servers and efficiently reduce the load of storage servers. Therefore, the scalability of the whole system is significantly enhanced. The experiment result shows that our solution can enhance the scalability of the Blue Whale System by 80% at most. Moreover, we adopt some optimizations to improve system performance further.