Symbiotic jobscheduling with priorities for a simultaneous multithreading processor

Abstract

Simultaneous Multithreading machines benefit from jobscheduling software that monitors how well coscheduled jobs share CPU resources, and coschedules jobs that interact well to make more efficient use of those resources. As a result, informed coscheduling can yield significant performance gains over naive schedulers. However, prior work on coscheduling focused on equal-priority job mixes, which is an unrealistic assumption for modern operating systems.This paper demonstrates that a scheduler for an SMT machine can both satisfy process priorities and symbiotically schedule low and high priority threads to increase system throughput. Naive priority schedulers dedicate the machine to high priority jobs to meet priority goals, and as a result decrease opportunities for increased performance from multithreading and coscheduling. More informed schedulers, however, can dynamically monitor the progress and resource utilization of jobs on the machine, and dynamically adjust the degree of multithreading to improve performance while still meeting priority goals.Using detailed simulation of an SMT architecture, we introduce and evaluate a series of five software and hardware-assisted priority schedulers. Overall, our results indicate that coscheduling priority jobs can significantly increase system throughput by as much as 40%, and that (1) the benefit depends upon the relative priority of the coscheduled jobs, and (2) more sophisticated schedulers are more effective when the differences in priorities are greatest. We show that our priority schedulers can decrease average turnaround times for a random jobmix by as much as 33%.

Symbiotic Space-Sharing on SDSC’s DataStar System

“…While sampling, the scheduler would exhibit a configurable bias towards choosing combinations known to be more efficient. This approach has been shown effective in multithreading scenarios [23,24].…”

“…Symbiotic job scheduling was originally proposed for machines utilizing Simultaneous Multithreading [23,24], later known as Hyperthreading, and was subsequently refined by McGregor et. al.…”

Reducing Shared Cache Misses via dynamic Grouping and Scheduling on Multicores

“…Similarly, symbiosis is applied on the scheduling of tasks, hence symbiosis is a factor that indicates the performance of tasks that are co-scheduled and compete in hardware resources every cycle [15]. This factor may be based on system performance, system utilization, energy delay product, cores energy, cores power, average normalized turnaround time (time between submitting a job to the system and its completion) ,cache sensitivity and cache intensity, or average shared cache miss rate [15][16][17][18].…”

“…This factor may be based on system performance, system utilization, energy delay product, cores energy, cores power, average normalized turnaround time (time between submitting a job to the system and its completion) ,cache sensitivity and cache intensity, or average shared cache miss rate [15][16][17][18]. Throughout this paper, we will consider the symbiosis factor as the miss rate for the second level L2 shared cache.…”

“…Sampling This technique is known as SOS (Sample, Optimize, Symbiosis) [15,17,19]. It is based on producing a profile for all the possible combinations for scheduling the tasks, then taking the schedule decision based on the highest symbiosis factor.…”