Home // ICNS 2013, The Ninth International Conference on Networking and Services // View article

Optimization of Server Locations in Server Migration Service

Authors:
Yukinobu Fukushima
Tutomu Murase
Tokumi Yokohira
Tatsuya Suda

Keywords: cloud computing, servermigration service, integer-programming model, penalty

Abstract:
In server migration service (SMS), a work place (WP) refers to a computer that runs a virtual machine, and a server refers to a virtual machine that runs a server-side application of a network application (NW-App). In SMS, WPs are deployed at various locations in a network, and servers may migrate between WPs towards the users of the NW-App to achieve better QoS for the users. In SMS, an SMS provider tries to provide an NW-App provider with a certain level of QoS that they agree upon, and if the SMS provider fails, it pays penalty (e.g., reimbursement of a part of service charges to users) depending on the degree and length of the QoS violation. Thus, the SMS provider is incentivized to migrate servers between WPs to satisfy the agreed-upon QoS level to reduce the penalty. On the other hand, an SMS provider also needs to be moderate in performing server migrations to avoid degradation of network QoS (i.e., QoS of background traffic). This is because a server is typically large in size and the server generates a large amount of traffic when it migrates, resulting in increasing delay and loss for its background traffic in a network. This paper formulates an integer-programming model for the off-line server locations decision (i.e., when and to which WP server should migrate) where the penalty associated with NW-App’s QoS violations is minimized, keeping the number and distance of server migrations below a given level. This paper also compares the minimum penalty obtained through solving the integer-programming model against the penalty obtained with a greedy on-line server locations decision algorithm, which migrates a server to a WP that minimizes the current penalty with no consideration of the penalty that will arise in the future. Numerical examples show that the integerprogramming model achieves 36% to 49% lower penalty than the greedy algorithm when the degradation of network QoS is little acceptable.

Pages: 200 to 206

Copyright: Copyright (c) IARIA, 2013

Publication date: March 24, 2013

Published in: conference

ISSN: 2308-4006

ISBN: 978-1-61208-256-1

Location: Lisbon, Portugal

Dates: from March 24, 2013 to March 29, 2013