.png){kind=icon}
.png){kind=icon}
Return
Cost-Optimized Server Allocation and Request Routing in Cloud Systems Using Different Service Classes
.png){kind=icon}
PRE
AI
DOI：10.1109/TCC.2026.3657060.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Cloud systems offer scalable and flexible infrastructure but can become expensive if the resource usage is not optimized. To help reduce service delivery costs, this paper presents cost-optimized load-balancing policies for several service classes and uses these to provide valuable system insights. Leveraging a comprehensive system model and considering diverse workload and system parameters, we explore the interplay between resource allocation, request routing, and system performance. Specifically, for each service class, our policies (1) determine the ideal server allocation strategy and (2) identify the most suitable replica site for each incoming request. By deriving optimized policies for different service models and conducting a comparative evaluation, we offer insights into optimal request routing strategies as well as when basic policies are optimal, perform close to optimal, or when more advanced policies are needed. Our findings contribute to improved load-balancing techniques in cloud computing, facilitating enhanced performance and efficiency in distributed environments.
Keywords:
Cloud computing
edge cloud
replica servers
optimization
server allocation
request routing
M/M/1
M/D/ $\infty$ ∞
Journal
I
IF:
5
Papers:
1.8K
Citations:
4.3K