Adaptive Optimization of a System's Load

摘要: Applications of modeling techniques based on queueing theory to computer system performance analysis normally assume the existence steady-state conditions. However, these conditions are often violated since unpredictable composition workload causes peaks having highly variable intensities and durations. Furthermore, is dependent how reacts fluctuations. Automatic control mechanisms required take care high variance resource demands. Real-time optimization overall a requires introduction adaptive controlled functions, An scheduling algorithm which controls input in order maximize given criterion, such as throughput, presented. The load adjusted depending characteristics both mix jobs execution submitted waiting queue. asymptotic bounds provides useful information about limits indexes that can be achieved with multiclass workload. evaluation performed through simulation experiments using data collected from two real workloads. This technique could used optimize throughput centralized well for automatic balancing distributed environment.