Scaling Into the Cloud

Date Added: Mar 2009
Format: PDF

As the use of cloud computing becomes more prevalent among popular web services, the issue arise of effectively scaling resources in the cloud. There are services that provide scaling automatically but require the use of specific platforms. The authors analyze three platform-agnostic algorithms for scaling resources dynamically: one developed by RightScale, and two others that predict system loads based on linear regression and autoregression of order 1. In order to effectively compare these algorithms, they developed a novel scoring metric based on availability and the standard cost model provided by cloud hosting services. The results show that dynamic provisioning provides marked improvements over static allocation in terms of cost with minimal drop in availability.