Optimized Hypervisor Scheduler for Parallel Discrete Event Simulations on Virtual Machine Platforms

Abstract

With the advent of virtual machine (VM)-based platforms for parallel computing, it is now possible to execute parallel discrete event simulations (PDES) over multiple virtual machines, in contrast to executing directly over hardware as is traditionally done over the past decades. While mature VM-based parallel systems now offer compelling benefits such as serviceability, dynamic reconfigurability and overall cost effectiveness, the runtime performance of parallel applications can be significantly affected. In particular, most VM-based platforms are optimized for general workloads, but PDES execution exhibits unique dynamics significantly different from other workloads. Here we first present results from experiments that highlight the gross deterioration of the runtime performance of VM-based PDES simulations when executed using traditional VM schedulers, quantitatively showing the bad scaling properties of the scheduler as the number of VMs is increased. We also present a new scheduler optimized specifically for PDES applications, and describe its design and implementation. Experimental results obtained from running PDES benchmarks over VMs show over an order of magnitude improvement in the run time of the PDES-optimized scheduler relative to the regular VM scheduler, with over 20x reduction in run time of simulations using up to 64 VMs. The observations and results are timely for emerging cloud/VM-based installations, highlighting the need for PDES-specific support.

[Pub 134]

http://simutools.org/2013

Kalyan Perumalla
Kalyan Perumalla

Kalyan Perumalla is Founder and President of Discrete Computing, Inc. He led advanced research and development at ORNL and holds senior faculty appointments at UTK, GT, and UNL.

Next
Previous

Related