TY - JOUR
T1 - Efficient prioritized service recovery using content-aware routing mechanism in web server cluster
AU - Kang, Euisuk
AU - Lee, Sookheon
AU - Park, Myong Soon
PY - 2005
Y1 - 2005
N2 - In the future, many more people will use web service in Ubiquitous environments. SMD-cluster, which is one of previous research, is a proposed scalable cluster system to meet the clients' requirement. SMD-cluster has a highly scalable architecture with multiple front-end nodes. However, SMD-cluster cannot provide continuous web service to clients, even though all real servers are alive when a front-end node fails. It causes considerable damage to the system, which provides mission-critical services such as E-business, Internet banking, and home shopping. To overcome this problem, we propose an efficient service recovery method in SMD-cluster. Service recovery information is maintained at each real server to minimize the overhead on front-end nodes. Waste of network usage is minimized by reducing communication messages among nodes in a cluster. We propose fast service recovery using a contentaware routing mechanism, which considers the specific content type of a request, to give priority to mission critical services.
AB - In the future, many more people will use web service in Ubiquitous environments. SMD-cluster, which is one of previous research, is a proposed scalable cluster system to meet the clients' requirement. SMD-cluster has a highly scalable architecture with multiple front-end nodes. However, SMD-cluster cannot provide continuous web service to clients, even though all real servers are alive when a front-end node fails. It causes considerable damage to the system, which provides mission-critical services such as E-business, Internet banking, and home shopping. To overcome this problem, we propose an efficient service recovery method in SMD-cluster. Service recovery information is maintained at each real server to minimize the overhead on front-end nodes. Waste of network usage is minimized by reducing communication messages among nodes in a cluster. We propose fast service recovery using a contentaware routing mechanism, which considers the specific content type of a request, to give priority to mission critical services.
UR - http://www.scopus.com/inward/record.url?scp=26844499545&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=26844499545&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-31957-3_36
DO - 10.1007/978-3-540-31957-3_36
M3 - Conference article
AN - SCOPUS:26844499545
SN - 0302-9743
VL - 3421
SP - 297
EP - 306
JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
IS - II
T2 - Networking - ICN 2005
Y2 - 17 April 2005 through 21 April 2005
ER -