The Computing Innovation Fellows Project

Posted on: Wednesday, May 12th, 2010
Broad Research Area: Networks / Operating Systems

Research Interests:

Cloud computing promises to reduce the cost of IT organizations by allowing them to purchase just as much compute and storage resources as needed, only when needed, and through lower
capital and operational expense stemming from the cloud’s economies of scale. Yet, little is known about the performance of applications in the cloud, the response time variation induced
by network latency, and the scale of applications suited for deployment. Migrating enterprise networks to the cloud is viewed
as one of the major challenges.

The goal of our research is to develop frameworks and methodologies that enable application architects to (i) predict the performance of their applications when deployed in the cloud;
and (ii) plan how the applications should be architected so as to meet the target goals for performance. A particular focus of our work is enterprise applications, which are not only large-scale with complex component interdependencies, but also involve stringent requirements on availability and service latency. A major thrust of
our research is in characterizing real enterprise applications, as there is little understanding of their characteristics.

We are conducting research along various dimensions such as (i) models that estimate the benefits (or costs) of (re)deploying
services in the cloud, given performance and availability constraints; (ii) techniques for analyzing
and predicting the performance of cloud applications while accounting for network latency, throughput and demand variability; and (iii) algorithms to determine how best deploy a given
application among cloud and private compute resources.

We are also developing methodologies for assurable reconfiguration of security policies as enterprises are migrated to the cloud, and exploring ways to specify and configure QoS policies in cloud platforms.

Contact Information:

Email sanjay@purdue.edu