Fiche de WLADDIMIRO Daniel

Titre : Dynamic adaptation in Stream Processing System
Date : 8 janvier 2024
Heure: 14:00
Résumé : Stream Processing Systems (SPSs) dynamically process input events. Since the input is usually not a constant flow, presenting rate fluctuations, many works in the literature propose to dynamically replicate SPS operators, aiming at reducing the processing bottleneck induced by such fluctuations. We present in this these a two approach to dynamically changes the number of replicas of an operator. First, a reactive approach, which based multiple metrics. And second, a predictive model, which based on input rate variation, execution time of operators, and queued events, dynamically defines the necessary current number of replicas of each operator. We also propose a Storm-based SPS, which uses such our model, not requiring reboot reconfiguration when the number of operators replica change.
Lieu : LIP6, Sorbonne Université, Paris, France