Packet task processing in distributed computational environments with inalienable resources

We study the planning problems for a packet of independent tasks in a virtual user organization of a distributed computational environment with inalienable resources. The plan is constructed based on a dynamically updated list of slots, each of which corresponds to a time interval during which a certain kind of resource is available. The planning is done cyclically, and the set of available slots is known at the beginning of every planning cycle. We propose a scheme to find the necessary number of slots suitable for completing the task. We consider the solutions of search problems for an optimal, in the single-criterion setting, and efficient, for a given vector of criteria, slot combinations with dynamic programming.