Scheduling best-effort and real-time pipelined applications on time-shared clusters

Research output: Contribution to conferencePaper

17 Citations (Scopus)

Abstract

Two important emerging trends are influencing the design, implementation and deployment of high performance parallel systems. The first is on the architectural end, where both economic and technological factors are compelling the use of off-the-shelf computing elements (workstations/PCs and networks) to put together high performance systems called clusters. The second is from the user community that is finding an increasing number of applications to benefit from such high performance systems. Apart from the scientific applications that have traditionally needed supercomputing power, a large number of graphics, visualization, database, web service and e-commerce applications have started using clusters because of their high processing and storage requirements. These applications have diverse characteristics and can place different Quality-of-Service (QoS) requirements on the underlying system (low response time, high throughput, high I/O demands, guaranteed response/throughput etc.). Further, clusters run ning such applications need to cater to potentially a large number of users (or other applications) in a time-shared manner. The underlying system needs to accommodate the requirements of each application, while ensuring that they do not interfere with each other. This paper focuses on the CPU resources of a cluster and investigates scheduling mechanisms to meet the responsiveness, throughput and guaranteed service requirements of different applications. Specifically, we propose and evaluate three different scheduling mechanisms. These mechanisms have been drawn from traditional solutions on parallel systems (gang scheduling and dynamic coscheduling), and have been extended to accommodate the new criteria under consideration. These mechanisms have been investigated using detailed simulation and workload models to show their pros and cons for different performance metrics.

Original languageEnglish (US)
Pages209-218
Number of pages10
StatePublished - Jan 1 2001
Event13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001) - Crete Island, Greece
Duration: Jul 3 2001Jul 6 2001

Other

Other13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001)
CountryGreece
CityCrete Island
Period7/3/017/6/01

Fingerprint

Scheduling
Throughput
Web services
Program processors
Quality of service
Visualization
Economics
Processing

All Science Journal Classification (ASJC) codes

  • Software
  • Safety, Risk, Reliability and Quality

Cite this

Zhang, Y., & Sivasubramaniam, A. (2001). Scheduling best-effort and real-time pipelined applications on time-shared clusters. 209-218. Paper presented at 13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001), Crete Island, Greece.
Zhang, Y. ; Sivasubramaniam, Anand. / Scheduling best-effort and real-time pipelined applications on time-shared clusters. Paper presented at 13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001), Crete Island, Greece.10 p.
@conference{501e5b55c16a4767ab182544ba194396,
title = "Scheduling best-effort and real-time pipelined applications on time-shared clusters",
abstract = "Two important emerging trends are influencing the design, implementation and deployment of high performance parallel systems. The first is on the architectural end, where both economic and technological factors are compelling the use of off-the-shelf computing elements (workstations/PCs and networks) to put together high performance systems called clusters. The second is from the user community that is finding an increasing number of applications to benefit from such high performance systems. Apart from the scientific applications that have traditionally needed supercomputing power, a large number of graphics, visualization, database, web service and e-commerce applications have started using clusters because of their high processing and storage requirements. These applications have diverse characteristics and can place different Quality-of-Service (QoS) requirements on the underlying system (low response time, high throughput, high I/O demands, guaranteed response/throughput etc.). Further, clusters run ning such applications need to cater to potentially a large number of users (or other applications) in a time-shared manner. The underlying system needs to accommodate the requirements of each application, while ensuring that they do not interfere with each other. This paper focuses on the CPU resources of a cluster and investigates scheduling mechanisms to meet the responsiveness, throughput and guaranteed service requirements of different applications. Specifically, we propose and evaluate three different scheduling mechanisms. These mechanisms have been drawn from traditional solutions on parallel systems (gang scheduling and dynamic coscheduling), and have been extended to accommodate the new criteria under consideration. These mechanisms have been investigated using detailed simulation and workload models to show their pros and cons for different performance metrics.",
author = "Y. Zhang and Anand Sivasubramaniam",
year = "2001",
month = "1",
day = "1",
language = "English (US)",
pages = "209--218",
note = "13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001) ; Conference date: 03-07-2001 Through 06-07-2001",

}

Zhang, Y & Sivasubramaniam, A 2001, 'Scheduling best-effort and real-time pipelined applications on time-shared clusters' Paper presented at 13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001), Crete Island, Greece, 7/3/01 - 7/6/01, pp. 209-218.

Scheduling best-effort and real-time pipelined applications on time-shared clusters. / Zhang, Y.; Sivasubramaniam, Anand.

2001. 209-218 Paper presented at 13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001), Crete Island, Greece.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Scheduling best-effort and real-time pipelined applications on time-shared clusters

AU - Zhang, Y.

AU - Sivasubramaniam, Anand

PY - 2001/1/1

Y1 - 2001/1/1

N2 - Two important emerging trends are influencing the design, implementation and deployment of high performance parallel systems. The first is on the architectural end, where both economic and technological factors are compelling the use of off-the-shelf computing elements (workstations/PCs and networks) to put together high performance systems called clusters. The second is from the user community that is finding an increasing number of applications to benefit from such high performance systems. Apart from the scientific applications that have traditionally needed supercomputing power, a large number of graphics, visualization, database, web service and e-commerce applications have started using clusters because of their high processing and storage requirements. These applications have diverse characteristics and can place different Quality-of-Service (QoS) requirements on the underlying system (low response time, high throughput, high I/O demands, guaranteed response/throughput etc.). Further, clusters run ning such applications need to cater to potentially a large number of users (or other applications) in a time-shared manner. The underlying system needs to accommodate the requirements of each application, while ensuring that they do not interfere with each other. This paper focuses on the CPU resources of a cluster and investigates scheduling mechanisms to meet the responsiveness, throughput and guaranteed service requirements of different applications. Specifically, we propose and evaluate three different scheduling mechanisms. These mechanisms have been drawn from traditional solutions on parallel systems (gang scheduling and dynamic coscheduling), and have been extended to accommodate the new criteria under consideration. These mechanisms have been investigated using detailed simulation and workload models to show their pros and cons for different performance metrics.

AB - Two important emerging trends are influencing the design, implementation and deployment of high performance parallel systems. The first is on the architectural end, where both economic and technological factors are compelling the use of off-the-shelf computing elements (workstations/PCs and networks) to put together high performance systems called clusters. The second is from the user community that is finding an increasing number of applications to benefit from such high performance systems. Apart from the scientific applications that have traditionally needed supercomputing power, a large number of graphics, visualization, database, web service and e-commerce applications have started using clusters because of their high processing and storage requirements. These applications have diverse characteristics and can place different Quality-of-Service (QoS) requirements on the underlying system (low response time, high throughput, high I/O demands, guaranteed response/throughput etc.). Further, clusters run ning such applications need to cater to potentially a large number of users (or other applications) in a time-shared manner. The underlying system needs to accommodate the requirements of each application, while ensuring that they do not interfere with each other. This paper focuses on the CPU resources of a cluster and investigates scheduling mechanisms to meet the responsiveness, throughput and guaranteed service requirements of different applications. Specifically, we propose and evaluate three different scheduling mechanisms. These mechanisms have been drawn from traditional solutions on parallel systems (gang scheduling and dynamic coscheduling), and have been extended to accommodate the new criteria under consideration. These mechanisms have been investigated using detailed simulation and workload models to show their pros and cons for different performance metrics.

UR - http://www.scopus.com/inward/record.url?scp=0034825294&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034825294&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0034825294

SP - 209

EP - 218

ER -

Zhang Y, Sivasubramaniam A. Scheduling best-effort and real-time pipelined applications on time-shared clusters. 2001. Paper presented at 13th Annual Symposium on Parallel Algorithms and Architectures (SPAA 2001), Crete Island, Greece.