Towards combining reactive and proactive cloud elasticity on running HPC applications

The elasticity feature of cloud computing has been proved as pertinent for parallel applications, since users do not need to take care about the best choice for the number of processes/resources beforehand. To accomplish this, the most common approaches use threshold-based reactive elasticity or time-consuming proactive elasticity. However, both present at least one problem related to: the need of a previous user experience, lack on handling load peaks, completion of parameters or design for a specific infrastructure and workload setting. In this regard, we developed a hybrid elasticity service for parallel applications named SelfElastic. As parameterless model, SelfElastic presents a closed control loop elasticity architecture that adapts at runtime the values of lower and upper thresholds. Besides presenting SelfElastic, our purpose is to provide a comparison with our previous work on reactive elasticity called AutoElastic. The results present the SelfElastic’s lightweight feature, besides highlighting its performance competitiveness in terms of application time and cost metrics.