Run-time Support for Scheduling Parallel Applications in Heterogeneous NOWs

نویسندگان

  • Jon B. Weissman
  • Xin Zhao
چکیده

This paper describes the current state of Prophet — a system that provides run-time scheduling support for parallel applications in heterogeneous workstation networks. Prior work on Prophet [10][12] demonstrated that scheduling SPMD applications could be effectively automated with excellent performance. Enhancements have been made to Prophet to broaden its use to other application types including parallel pipelines, and to make more effective use of dynamic system state information to further improve performance. The results indicate that both SPMD and parallel pipeline applications can be scheduled to produce reduced completion time by exploiting the application structure and run-time information.1

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

EXPLORER: Supporting Run-Time Parallelization of DO-ACROSS Loops on General Networks of Workstations

Performing runtime parallelization on general networks of workstations (NOWs) without special hardware or system software supports is very diicult, especially for DOACROSS loops. With the high communication overhead on NOWs, there is hardly any performance gain for runtime parallelization, due to the latter's large amount of messages for dependence detection, data accesses, and computation sche...

متن کامل

QUIC: A Quality of Service Network Interface Layer for Communication in NOWs

This project explores the development of a hard-ware/software infrastructure to enable the provision of quality of service (QoS) guarantees in high performance networks used to conngure clusters of worksta-tions/PCs. These networks of workstations (NOWs) have emerged as a viable high performance computational vehicle and are also being called upon to support access to multimedia datasets. Examp...

متن کامل

Load Balancing for Minimizing Execution Time of a Target Job on a Network of Heterogeneous Workstations

A network of workstations (NOWs) may be employed for high performance computing where execution time of a target job is to be minimized. Job arrival rate and size are “random” on a NOWs. In such an environment, partitioning (load balancing) a target job based on only the first order moments (means) of system parameters is not optimal. In this paper, it is proposed to consider the second order m...

متن کامل

Solving the Problem of Scheduling Unrelated Parallel Machines with Limited Access to Jobs

Nowadays, by successful application of on time production concept in other concepts like production management and storage, the need to complete the processing of jobs in their delivery time is considered a key issue in industrial environments. Unrelated parallel machines scheduling is a general mood of classic problems of parallel machines. In some of the applications of unrelated parallel mac...

متن کامل

Solving the Problem of Scheduling Unrelated Parallel Machines with Limited Access to Jobs

Nowadays, by successful application of on time production concept in other concepts like production management and storage, the need to complete the processing of jobs in their delivery time is considered a key issue in industrial environments. Unrelated parallel machines scheduling is a general mood of classic problems of parallel machines. In some of the applications of unrelated parallel mac...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1997