Evaluation of Design Choices for Gang Scheduling Using Distributed Hierarchical Control

Feitelson, Dror G.; Rudolph, Larry

doi:10.1006/jpdc.1996.0064

Cited by 49 publications

(25 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The DHC scheme has been evaluated with three different distributions: a uniform distribution in which all sizes are equally likely, a harmonic distribution in which the probability of size s is proportional to 1/s, and a uniform distribution on powers of two. Both analysis and simulations showed significant differences between the utilizations that could be achieved for the three distributions (Figure 1.1) [245]. These differences corresponds to different degrees of fragmentation that are inherent to packing jobs that come from these distributions.…”

Section: Example 2: Processor Allocation Using a Buddy Systemmentioning

confidence: 93%

“…One example is provided by the DHC scheme [245], in which a buddy system is used for processor Simulation results showing normalized response time (slowdown) as a function of load for processor allocation to parallel jobs using DHC, from [245]. The three curves are for exactly the same system -the only difference is in the distribution of job sizes.…”

Section: Example 2: Processor Allocation Using a Buddy Systemmentioning

confidence: 99%

See 1 more Smart Citation

Workload Modeling for Computer Systems Performance Evaluation

Feitelson

2015

Self Cite

197

131

View full text Add to dashboard Cite

Reliable performance evaluations require the use of representative workloads. This is no easy task since modern computer systems and their workloads are complex, with many interrelated attributes and complicated structures. Experts often use sophisticated mathematics to analyze and describe workload models, making these models difficult for practitioners to grasp. This book aims to close this gap by emphasizing the intuition and the reasoning behind the definitions and derivations related to the workload models. It provides numerous examples from real production systems, with hundreds of graphs. Using this book, readers will be able to analyze collected workload data and clean it if necessary, derive statistical models that include skewed marginal distributions and correlations, and consider the need for generative models and feedback from the system. The descriptive statistics techniques covered are also useful for other domains.

show abstract

Section: Example 2: Processor Allocation Using a Buddy Systemmentioning

confidence: 93%

Section: Example 2: Processor Allocation Using a Buddy Systemmentioning

confidence: 99%

Workload Modeling for Computer Systems Performance Evaluation

Feitelson

2015

Self Cite

197

131

View full text Add to dashboard Cite

show abstract

“…To schedule more than P processes, the system is time-shared. We will assume processes are single-threaded, and all P processors context switch at the same time as would be done in gang scheduling [5]. These assumptions are not central to our approach, rather for the sake of brevity, we have focused on a basic scheduling scenario.…”

Section: Scheduling Problemmentioning

confidence: 99%

A new memory monitoring scheme for memory-aware scheduling and partitioning

Suh¹,

Devadas²,

Rudolph³

Proceedings Eighth International Symposium on High Performance Computer Architecture

192

174

View full text Add to dashboard Cite

show abstract

“…Generally speaking, parallel job schedulers have been designed to deal with CPU-intensive jobs [4,6,7,8,13,14,15,16,20,23,25]. Some researches have proposed strategies to deal with I/O-intensive and irregular jobs [2,3,17,18,19,22,24], but not with I 3 jobs.…”

Section: Introductionmentioning

confidence: 99%