Trace-Driven Simulation for Energy Consumption in High Throughput Computing Systems

Abstract

any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Additional information:Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. Abstract-High Throughput Computing (HTC) is a powerful paradigm allowing vast quantities of independent work to be performed simultaneously. However, until recently little evaluation has been performed on the energy impact of HTC. Many organisations now seek to minimise energy consumption across their IT infrastructure though it is unclear how this will affect the usability of HTC systems. We present here HTC-Sim, a simulation system which allows the evaluation of different energy reduction policies across an HTC system comprising a collection of computational resources dedicated to HTC work and resources provided through cycle scavenging -a Desktop Grid. We demonstrate that our simulation software scales linearly with increasing HTC workload.

Energy-Aware Simulation of Workflow Execution in High Throughput Computing Systems

“…We previously introduced HTC-Sim [3], a Java-based tracedriven simulation environment, demonstrating novelty in its ability to model the scheduling decisions of a HTC system in multi-use cluster environments, and the inclusion of faulttolerance mechanisms [23]. In this work we extend HTCSim to support workflow execution and scheduling with an emphasis on energy consumption.…”

“…As jobs may be evicted and re-run elsewhere this leads to an increase in their makespan (time between job submission and final results becoming available) and energy consumption -which can be detrimental to both the job submitter and infrastructure owner. In previous work we have evaluated these impacts and shown how we can reduce the energy consumption without adversely affecting individual job makespan [3].…”

“…In this work we extend our earlier HTC simulation system (HTC-Sim [3]) to allow for the execution of workflows on a non-dedicated hardware infrastructure. This simulation allows us to rapidly evaluate the impact on high throughput users in terms of the impact of running their jobs and workflows, but also on the energy impact that this has on the overall system.…”

“…analysis, to determine how possible configuration changes could affect the system more rapidly than deploying to a real environment. Major inroads have been made in the areas of individual system energy consumption [7] and more recently in the energy consumption of collections of computer systems -this has included Cloud simulations [8], simulations of energy consumption within non-dedicated volunteer computing systems for HTC [3] such as the scenario outlined in Section VI-B, and simulations of energy consumption of workflows running on dedicated hardware [9]. However, to the best of our knowledge, noone has attempted to model the energy impact of workflows running over a non-dedicated volunteer computing cluster.…”

HTC‐Sim: a trace‐driven simulation framework for energy consumption in high‐throughput computing systems

“…This policy was first introduced in [70] as an extension of resource allocation strategies first explored in [56].…”