log2cloud: log-based prediction of cost-performance trade-offs for cloud deployments

Authors:
Diego Perez-Palacin;Radu Calinescu;José Merseguer
Affiliations:
Universidad de Zaragoza, Spain;University of York, United Kingdom;Universidad de Zaragoza, Spain
Venue:
Proceedings of the 28th Annual ACM Symposium on Applied Computing
Year:
2013

Citing 10
Cited 0

Analysis of Log Files Intersections for Security Enhancement

ITNG '06 Proceedings of the Third International Conference on Information Technology: New Generations
A break in the clouds: towards a cloud definition

ACM SIGCOMM Computer Communication Review
Cloud computing and emerging IT platforms: Vision, hype, and reality for delivering computing as the 5th utility

Future Generation Computer Systems
The Real Cost of a CPU Hour

Computer
On Technical Security Issues in Cloud Computing

CLOUD '09 Proceedings of the 2009 IEEE International Conference on Cloud Computing
A view of cloud computing

Communications of the ACM
Cloud Migration: A Case Study of Migrating an Enterprise IT System to IaaS

CLOUD '10 Proceedings of the 2010 IEEE 3rd International Conference on Cloud Computing
Advances and challenges in log analysis

Communications of the ACM
The Cloud Adoption Toolkit: supporting cloud adoption decisions in the enterprise

Software—Practice & Experience
Specification and quantitative analysis of probabilistic cloud deployment patterns

HVC'11 Proceedings of the 7th international Haifa Verification conference on Hardware and Software: verification and testing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Numerous organisations are considering moving at least some of their existing applications to the cloud. A key motivating factor for this fast-paced adoption of cloud is the expectation of cost savings. Estimating what these cost savings might be requires comparing the known cost of running an application in-house with a predicted cost of its cloud deployment. A major problem with this approach is the lack of suitable techniques for predicting the cost of the virtual machines (VMs) that a cloud-deployed application requires in order to achieve a given service-level agreement. We introduce a technique that addresses this problem by using established results from queueing network theory to predict the minimum VM cost of cloud deployments starting from existing application logs. We describe how this formal technique can be used to predict the cost-performance trade-offs available for the cloud deployment of an application, and presents a case study based on a real-world webmail service.