Middleware & System Software

PC˛'s research focuses on the problem of how to guarantee the use of the resources with Service Level Agreements (SLA). This research includes the ability to perform fault tolerance mechanisms like checkpointing and migration of SLA agreed jobs and research on the assessment about likelihood of SLA violations. The combined instruments, risk assessment and fault tolerance mechanisms, allow a powerful risk aware management of grid/cloud  jobs. This improves the guaranteed service quality of the resource management. In addition the PC˛ worked on the integration of Web service based Enterprise Application Integration (EAI) into the Grid. Our aim is to combine the strengths of the two areas, loosely coupled services and the secure and easy to deploy grid/cloud infrastructures. The result will be the ability to create secure business workflows in these infrastructurdes. These fields are vital requirements for future commercial use of grid/cloud environments.

Resource Management Systems (RMS) are needed for the cloud as well as for compute clusters. They allow users and system administrators to access and manage various computing resources like processors, memory, networks, or storage. PC˛ has developed an expandable and modular RMS, called Computing Center Software, which uses a planning based job scheduler. This OpenCCS is used in several projects and its features are continuously extended.

Projects, which belong to this research subjects, are:

  • Hydra: Network Embedded System Middleware
  • MoSGrid: Molecular Simulation Grid
  • OpenCCS: Computing Center Software, a resource management software for HPC systems
  • RECS: Resource Efficient Cluster System
  • BIS-Grid: Grid-based integration and orchestration of business information systems
  • DGSI: D-Grid Scheduler Interoperability
  • EDGI: European Desktop Grid Initiative
Research