Improving the Scalability of Parallel Jobs by adding Parallel Awareness to the Operating System

A parallel application benefits from scheduling policies that include a global perspective of the application's process working set. As the interactions among cooperating processes increase, mechanisms to ameliorate waiting within one or more of the processes become more important. In particula...

Full description

Saved in:
Bibliographic Details
Main Authors: Jones, Terry, Dawson, Shawn, Neely, Rob, Tuel, William, Brenner, Larry, Fier, Jeffrey, Blackmore, Robert, Caffrey, Patrick, Maskell, Brian, Tomlinson, Paul, Roberts, Mark
Format: Conference Proceeding
Language:English
Subjects:
Application software
Computing methodologies
General and reference
Kernel
Laboratories
Linux
Operating systems
Processor scheduling
Scalability
Software and its engineering
Software and its engineering > Software organization and properties
Software and its engineering > Software organization and properties > Contextual software domains
Software and its engineering > Software organization and properties > Contextual software domains > Operating systems
Software and its engineering > Software organization and properties > Contextual software domains > Operating systems > Process management
Software and its engineering > Software organization and properties > Contextual software domains > Operating systems > Process management > Scheduling
Switches
Theory of computation
Theory of computation > Design and analysis of algorithms
Theory of computation > Design and analysis of algorithms > Approximation algorithms analysis
Theory of computation > Design and analysis of algorithms > Approximation algorithms analysis > Scheduling algorithms
Theory of computation > Design and analysis of algorithms > Online algorithms
Theory of computation > Design and analysis of algorithms > Online algorithms > Online learning algorithms
Theory of computation > Design and analysis of algorithms > Online algorithms > Online learning algorithms > Scheduling algorithms
USA Councils
Weapons
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A parallel application benefits from scheduling policies that include a global perspective of the application's process working set. As the interactions among cooperating processes increase, mechanisms to ameliorate waiting within one or more of the processes become more important. In particular, collective operations such as barriers and reductions are extremely sensitive to even usually harmless events such as context switches among members of the process working set. For the last 18 months, we have been researching the impact of random short-lived interruptions such as timer-decrement processing and periodic daemon activity, and developing strategies to minimize their impact on large processor-count SPMD bulk-synchronous programming styles. We present a novel co-scheduling scheme for improving performance of fine-grain collective activities such as barriers and reductions, describe an implementation consisting of operating system kernel modifications and run-time system, and present a set of empirical results comparing the technique with traditional operating system scheduling. Our results indicate a speedup of over 300% on synchronizing collectives.
DOI:10.1145/1048935.1050161