TY - GEN
T1 - DISH - Dynamic information-based scalable hashing on a cluster of web cache servers
AU - Sohn, Andrew
AU - Kwak, Hukeun
AU - Chung, Kyusik
PY - 2007
Y1 - 2007
N2 - Caching web pages is an important part of web infrastructure. The effects of caching services are even more pronounced for wireless infrastructures due to their limited bandwidth. Medium to large-scale infrastructures deploy a cluster of servers to solve the scalability problem and hot spot problem inherent in caching. In this report, we present Dynamic Information-based Scalable Hashing (DISH) that evenly hashes client requests to a cluster of cache servers. Three types of runtime information are used to determine when and how to cache pages, including cache utilization, CPU usage, and number of connections. Pages cached are stored and retrieved mutually exclusively to/from all the servers. We have implemented our approach and performed various experiments using publicly available traces. Experimental results on a cluster of 16 cache servers demonstrate that the proposed hashing method gives 45% to 114% performance improvement over other widely used methods, while addressing the hot spot problem.
AB - Caching web pages is an important part of web infrastructure. The effects of caching services are even more pronounced for wireless infrastructures due to their limited bandwidth. Medium to large-scale infrastructures deploy a cluster of servers to solve the scalability problem and hot spot problem inherent in caching. In this report, we present Dynamic Information-based Scalable Hashing (DISH) that evenly hashes client requests to a cluster of cache servers. Three types of runtime information are used to determine when and how to cache pages, including cache utilization, CPU usage, and number of connections. Pages cached are stored and retrieved mutually exclusively to/from all the servers. We have implemented our approach and performed various experiments using publicly available traces. Experimental results on a cluster of 16 cache servers demonstrate that the proposed hashing method gives 45% to 114% performance improvement over other widely used methods, while addressing the hot spot problem.
UR - http://www.scopus.com/inward/record.url?scp=38149004370&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=38149004370&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-75444-2_73
DO - 10.1007/978-3-540-75444-2_73
M3 - Conference contribution
AN - SCOPUS:38149004370
SN - 9783540754435
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 785
EP - 796
BT - High Performance Computing and Communications - Third International Conference, HPCC 2007, Proceedings
PB - Springer Verlag
T2 - 3rd International Conference on High Performance Computing and Communications, HPCC 2007
Y2 - 26 September 2007 through 28 September 2007
ER -