Section: New Software and Platforms

Infrastructure Middleware and scheduler

Participant : Olivier Richard [correspondent] .

OAR

– The OAR project (see http://oar.imag.fr ) focuses on robust and highly scalable batch scheduling for clusters and grids. Its main objectives are the validation of grid administration tools such as Taktuk , the development of new paradigms for grid scheduling and the experimentation of various scheduling algorithms and policies.

The grid development of OAR has already started with the integration of best effort jobs whose purpose is to take advantage of idle times of the resources. Managing such jobs requires a support of the whole system from the highest level (the scheduler has to know which tasks can be canceled) down to the lowest level (the execution layer has to be able to cancel awkward jobs). OAR is perfectly suited to such developments thanks to its highly modular architecture. Moreover, this development is used for the CiGri grid middleware project.

The OAR system can also be viewed as a platform for the experimentation of new scheduling algorithms. Current developments focus on the integration of theoretical batch scheduling results into the system so that they can be validated experimentally.

CiGri

(http://cigri.imag.fr/ ) is a middleware which gathers the unused computing resource from intranet infrastructure and makes it available for the processing of large set of tasks. It manages the execution of large sets of parametric tasks on lightweight grid by submitting individual jobs to each batch scheduler. It is s associated to the OAR resource management system (batch scheduler). Users can easily monitor and control their set of jobs through a web portal. CiGri provides mechanisms to identify job error causes, to isolate faulty components and to resubmit jobs in a safer context.

ComputeMode

(http://computemode.imag.fr/ ) is an software infrastructure that allows to extends or create a Grid through the aggregation of unused computing resources. For instance, a virtual cluster can be built using anyone's PC while not in use. Indeed, most PCs in large companies or university campus are idle at night, on weekends, and during vacations, training periods or business trips.