Operation of an SLA-aware Grid Fabric

Research on Grid Computing started under purely technical questions of how to realize the virtualization of distributed resources. Now, the commercial user should be attracted to use the Grid. Future applications of the Next Generation Grid will demand for flexible negotiation mechanisms supporting various ways of Quality-of-Service (QoS) guarantees. In this context a Service Level Agreement (SLA) is a powerful instrument for describing all obligations and expectations within a business partnership. Many research projects already focus on realizing SLAs within the Grid middleware. However this is not sufficient. Resource Management Systems also have to be SLA-aware, since these systems provide their resources to Grid infrastructures. In this study we present the EU-funded project HPC4U (Highly Predictable Clusters for Internet Grids), which aims at realizing such an RMS by means of SLA-negotiation and scheduling and application-transparent fault tolerance.