. Jobs that do not require all processors in the system can be packed together for gang scheduling. We examine accounting traces from several parallel computers to show that indeed many jobs have small sizes and can be packed together. We then formulate a number of such packing algorithms, and evaluate their effectiveness using simulations based on our workload study. The results are that two algorithms are the best: either perform the mapping based on a buddy system of processors, or use migration to re-map the jobs more tightly whenever a job arrives or terminates. Other approaches, such as mapping to the least loaded PEs, proved to be counterproductive. The buddy system approach depends on the capability to gang-schedule jobs in multiple slots, if there is space. The migration algorithm is more robust, but is expected to suffer greatly due to the overhead of the migration itself. In either case fragmentation is not an issue, and utilization may top 90% with sufficiently...

. Statistics of a parallel workload on a 128-node iPSC/860 located at NASA Ames are presented. It is shown that while the number of sequential jobs dominates the number of parallel jobs, most of the resources (measured in node-seconds) were consumed by parallel jobs. Moreover, most of the sequential jobs were for system administration. The average runtime of jobs grew with the number of nodes used, so the total resource requirements of large parallel jobs were larger by more than the number of nodes they used. The job submission rate during peak day activity was somewhat lower than one every two minutes, and the average job size was small. At night, submission rate was low but job sizes and system utilization were high, mainly due to NQS. Submission rate and utilization over the weekend were lower than on weekdays. The overall utilization was 50%, after accounting for downtime. About 2/3 of the applications were executed repeatedly, some for a significant number of times....

The performance potential of run-to-completion (RTC) parallel processor scheduling policies is investigated by examining whether (1) application execution rate characteristics such as average parallelism (avg) and processor working set (pws) and/or (2) limited preemption can be used to improve the performance of these policies. We address the first question by comparing policies (previous as well as new) that differ only in whether or not they use execution rate characteristics and by examining a wider range of the workload parameter space than previous studies. We address the second question by comparing a simple two-level queueing policy with RTC scheduling in the second level queue against RTC policies that don't allow any preemption and against dynamic equiallocation (EQ). Using simulation to estimate mean response times we find that for promising RTC policies such as adaptive static partitioning (ASP) and shortest demand first (SDF), a maximum allocation constraint that is for all...

Scheduling algorithms that use application and system knowledge have been shown to be more effective at scheduling parallel jobs on a multiprocessor than algorithms that do not. This paper focuses on obtaining such information for use by a scheduler in a network of workstations environment. The log files from three parallel systems are examined to determine both how to categorize parallel jobs for storage in a job database and what job information would be useful to a scheduler. A Historical Profiler is proposed that stores information about programs and users, and manipulates this information to provide schedulers with execution time predictions. Several preemptive and non-preemptive versions of the FCFS, EASY and Least Work First scheduling algorithms are compared to evaluate the utility of the profiler. It is found that both preemption and the use of application execution time predictions obtained from the Historical Profiler lead to improved performance.

When a moldable job is submitted to a space-sharing parallel computer, it must choose whether to begin execution on a small, available cluster or wait in queue for more processors to become available. To make

Abstract not found

When multiple jobs compete for processing resources on a parallel computer, the operating system kernel's processor allocation policy determines how many and which processors to allocate to each. This dissertation investigates the issues involved in constructing a processor allocation policy for large scale, message-passing parallel computers supporting a scientific workload. First, the issues that affect the performance of scheduling policies for message-passing parallel systems are examined. We argue that reasonable policies must provide nearly equal resource allocation to all runnable jobs and allocate, to a single job, processors that are in close proximity to one another. Second, the concept of efficiency preservation is defined as a characteristic of processor allocation policie...

Various techniques for multiprogramming parallel multiprocessor systems have been proposed recently as a way to improve performance. A natural approach is to divide the set of processing elements into independent partitions, and simultaneously execute a different parallel program in each partition. Several issues arise, including the determination of the optimal number of programs allowed to execute simultaneously (i.e., the number of partitions) and the corresponding partition sizes. This can be done statically, dynamically, or adaptively, depending on the system and workload characteristics. In this paper several adaptive partitioning policies are evaluated. Their behavior, as well as the behavior of static policies, is investigated using real parallel programs. The policy applicability to actual systems is addressed, and implementation results of the proposed policies on an iPSC/2 hypercube system are reported. The concept of robustness (i.e., the ability to perform well on a wide r...

While the parallel use of many processors is a major attraction of scalable multiprocessors for large applications, another important feature of such machines is the large amount of physical memory they make available. Despite these resources, truly large applications may be limited not only by the amount of computation they require, but also by the amount of data they must operate on to solve problems of interest. In this paper we consider the problem of multiprocessor scheduling of jobs whose memory requirements place lower bounds on the fraction of the machine required in order to execute. We address three primary questions in this work: 1. How can a parallel machine be multiprogrammed with minimal overhead when jobs have minimum memory requirements? 2. To what extent does the inability of an application to repartition its workload during runtime affect the choice of processor allocation policy? 3. How rigid should the system be in attempting to provide equal resource allocation to ...

We describe a scheduler based on the microeconomic paradigm for scheduling on-line a set of parallel jobs in a multiprocessor system. In addition to increasing the system throughput and reducing the response time, we consider fairness in allocating system resources among the users, and provide the user with control over the relative performances of his jobs. Every user has a savings account in which he receives money at a constant rate. To run a job, the user creates an expense account for that job to which he transfers money from his savings account. The job uses the funds in its expense account to obtain the system resources it needs. The share of the system resources allocated to the user is directly related to the rate at which the user receives money; the rate at which the user transfers money into a job expense account controls the job's performance. We prove that starvation is not possible in our model. Simulation results show that our scheduler improves both sy...