Computational resource management systems

Industrial processes require sufficient computational resources and their high availability, along with right tools and practices. Often, the amount of computational resources is not the problem, but the easy and efficient utilisation of the resources. To tackle this challenge, a class of information systems, computational distributed resource management systems (DRMSs), has been developed. These systems aim to optimise the computational load bal-ancing and resource availability for the users letting the user to focus on the actual challenge in hands. The trend in this sector is from managing the resources of a single system, such as a computational server, towards distributed large computational resources in form of compu-tational grids and cloud computing. This report discusses different software tools to improve the usability and utilisation of computational resources. There are different tools to utilise office laptop and desktop computers idle resources, dedi-cated computer clusters, grids of heterogeneous computational hardware, and cloud compu-ting resources. Some DRMSs can deal with several use scenarios, but none of them can handle all the scenarios well, and thus there is multitude of different workable systems. In this report, most of the major computational DRMSs are discussed, and use experiences are covered for the selected tools: Grid Engine, SLURM, HTCondor, and Techila. Grid Engine and SLURM are designed for cluster computation, i.e. for systems from small to vast collec-tions of dedicated computers connected with a special network. HTCondor's and Techila's strengths are in workstation grids, i.e. utilisation of idle computational resources of the work-stations and laptops in a heterogeneous computer network or cloud computing resources.