Quality of experience in distributed databases

摘要: This paper proposes an approach to improve the level of Quality Experience (QoE) that distributed database systems provide. Quality is a measure users' satisfaction when using certain service or application. Therefore, main objective this provide mechanisms increase accessing systems. In traditional systems, users cannot specify execution-related constraints. Then, system evaluate if user expectations are satisfied and neither can take corrective actions necessary. In work, we present QoE-oriented (QoE-DDB). It allow Data Access Requirements (DARs) aims please by satisfying DARs they define. We define set types propose some SQL extensions enable requirements. Proposed include execution deadline priority, start finish times, data availability freshness degrees, disconnected mode. In our QoE-DDB, each user's command transformed into one more tasks executed services. Community modules local services negotiate Service Level Objectives (SLOs) for task, which improves system's dependability. both scheduling dynamic placement strategies. architecture strategies be used in wide range environments, from tightly-coupled homogeneous environments (e.g. composed off-the-shelf computers connected LAN) highly heterogeneous geographically where have degree autonomy. Traditional performance indicators throughput response time) not adequate QoE provides. also specialized Key Performance Indicators (KPIs) estimate provides. Finally, experimental results obtained through use benchmark queries together with prototype implements proposed In experiments, consider realistic scenarios compare their best-effort oriented counterparts. Obtained prove importance approach.