Section: New Results

Heterogeneous Resource Management

Participants : Eliya Buyukkaya, Djawida Dib, Eugen Feller, Christine Morin, Nikolaos Parlavantzas, Guillaume Pierre.

Cross-resource scheduling in heterogeneous cloud environments

Participants : Eliya Buyukkaya, Guillaume Pierre.

Allocating resources to applications in a heterogeneous cloud environment is harder than in a homogeneous environment. In a heterogeneous cloud some rare resources are more precious than others, and should be treated carefully to maximize their utilization. Similarly, applications may request groups of resources that exhibit certain inter-resource properties such as the available bandwidth between the assigned resources. We are currently investigating scheduling algorithms for handling such scenarios.

Maximizing private cloud provider profit in cloud bursting scenarios

Participants : Djawida Dib, Christine Morin, Nikolaos Parlavantzas.

Current PaaS offerings either provide no support for SLA guarantees or provide limited support targeting a restricted set of application types. To overcome this limitation, we have developed an open, cloud-bursting PaaS system, called Meryn, designed to be easily extensible to host new application types. The system integrates a decentralized optimization policy that maximises the PaaS provider profit, taking into account the payment of penalties incurred when quality guarantees are unsatisfied. The system was implemented and evaluated on the Grid5000 testbed using batch and MapReduce workloads. The results demonstrated the effectiveness of the policy in increasing provider profit [16] This work was part of Djawida Dib's PhD thesis [10] defended in July 2014.

Data life-cycle management in clouds

Participants : Eugen Feller, Christine Morin.

Infrastructure as a Service (IaaS) clouds provide a flexible environment where users can choose and control various aspects of the machines of interest. However, the flexibility of IaaS clouds presents unique challenges for storage and data management in these environments. Users use manual and/or ad-hoc methods to manage storage and data in these environments. FRIEDA is a Flexible Robust Intelligent Elastic Data Management framework that employs a range of data management strategies approaches in elastic environments. This year, our work carried out in the context of the DALHIS associate team (http://project.inria.fr/dalhis ), was focused on the extended design and evaluation of the FRIEDA data management system. FRIEDA was tested to work on Amazon EC2 resources. In addition, we layered a commandline utility atop FRIEDA that allows users to plug-in applications to run in FRIEDA. These tools have been adopted by the LBL-ATLAS group to run their experiments on Amazon [29] .