High-level Implementation and Evaluation of Cloud Resource Allocation Mechanism

With the rise of shared computing infrastructures such as clouds or clusters, the fair allocation of multiple resources rapidly gains relevance in communication systems. In particular, different resources like CPU, RAM, disk space, and bandwidth have to be shared among users with different demands, such that the overall outcome can be considered fair, i.e., shares with different amounts of inherently different resources have to be compared. Although some notions to quantify the fairness of multi-resource allocations were presented, only one allocation mechanism has recently been introduced leaving the general problem under-explored. In particular, the mentioned algorithm proves to have desirable properties (Pareto optimality, strategy proofness, sharing incentive, envy freeness) but only works in limited scenarios. To close the gap of widely applicable multi-resource allocation mechanisms for computing infrastructures a comprehensible fairness-metric and an according allocation mechanism was developed and formalized at the CSG.

The next step in the development of the allocation mechanism is its evaluation by means of simulations (not real world deployment). The assignment therefore first demands the student to implement the specified algorithm in a high-level programming language of the students choice. Subsequently user behavior has to be modelled algorithmically by the student. During this step the student is expected to help in the anticipation of strategic user behavior, i.e., can users influence the outcome of the mechanism by making false statements about their demands. Such manipulation attempts just as the overall performance of the algorithm will be evaluated with the help of the student.

In summary, logical comprehension as well as algorithmic thinking are expected to be strengths of the student.