Workload Modeling for Computer Systems Performance Evaluation

Feitelson, Dror G.

doi:10.1017/cbo9781139939690

Cited by 196 publications

(138 citation statements)

References 620 publications

(1,368 reference statements)

Supporting

Mentioning

131

Contrasting

Unclassified

Order By: Relevance

“…In our experiments, the search index is built from a dump of the English version of Wikipedia from July 2013. Query terms are chosen randomly, following a Zipfian distribution, which has been shown to model online search query distributions well [7,20]. masstree [35] is a fast, scalable in-memory key-value store written in C++.…”

Section: Tailbench Applicationsmentioning

confidence: 99%

Tailbench: a benchmark suite and evaluation methodology for latency-critical applications

Kasture

Sánchez

2016

2016 IEEE International Symposium on Workload Characterization (IISWC)

150

View full text Add to dashboard Cite

Abstract-Latency-critical applications, common in datacenters, must achieve small and predictable tail (e.g., 95 th or 99 th percentile) latencies. Their strict performance requirements limit utilization and efficiency in current datacenters. These problems have sparked research in hardware and software techniques that target tail latency. However, research in this area is hampered by the lack of a comprehensive suite of latency-critical benchmarks.We present TailBench, a benchmark suite and evaluation methodology that makes latency-critical workloads as easy to run and characterize as conventional, throughput-oriented ones. TailBench includes eight applications that span a wide range of latency requirements and domains, and a harness that implements a robust and statistically sound load-testing methodology. The modular design of the TailBench harness facilitates multiple load-testing scenarios, ranging from multi-node configurations that capture network overheads, to simplified single-node configurations that allow measuring tail latency in simulation. Validation results show that the simplified configurations are accurate for most applications. This flexibility enables rapid prototyping of hardware and software techniques for latency-critical workloads.

show abstract

Section: Tailbench Applicationsmentioning

confidence: 99%

Tailbench: a benchmark suite and evaluation methodology for latency-critical applications

Kasture

Sánchez

2016

2016 IEEE International Symposium on Workload Characterization (IISWC)

150

View full text Add to dashboard Cite

show abstract

“…The workload test denotes that any workload used in performance studies can be classified as either real or synthetic [18], [19]. Since there are two types of load, there is the possibility that the execution and application may be either centralized or distributed.…”

Section: The Problem Statementmentioning

confidence: 99%

“…Furthermore, metrics can be employed to quantify the difference between the characteristics of the workload which reaches the target system rather than the workload that was specified. The metrics can include the position, dispersion, central tendency, focus and skew, and correlation measured, as well as, the characterization by means of statistical distributions that best fits the data [18], [19].…”

Section: The Problem Statementmentioning

confidence: 99%

PEESOS-Cloud: A Workload-Aware Architecture for Performance Evaluation in Service-Oriented Systems

Ferreira

Nunes

Pereira

et al. 2016

2016 IEEE World Congress on Services (SERVICES)

View full text Add to dashboard Cite

Abstract-It is a challenging task to ensure quality in serviceoriented systems deployed in cloud computing owing to the dynamicity of its environment. Many approaches have been adopted to identify and evaluate bottlenecks and problems in performance. The most common scenario consists of distributed systems that use a workload capable of enabling clients to exploit the target system in different operational conditions. However, one requirement that tends to be overlooked is to determine how the workload is executed, as software and hardware faults can lead to its mischaracterization. In this paper, a number of problems in the workload generation have been identified and summarized. A new architecture, called PEESOS-Cloud, is proposed which allows these services to be evaluated as well as to improve the ability of the workload so that it conforms with its described characteristics. Experiments in a cloud environment were conducted to show how PEESOS-Cloud works and validate its capabilities. Our experiment also showed that the mischaracterization of the workload leads to poor results, whereas an workload-aware implementation leads to a better performance evaluation.

show abstract

“…This distribution was carried out through CDF where CDF refers to cumulative distribution function denoted by F(x).Cumulative density function is defined as the probability that a sample is smaller than or equal to some given value. Equation (1) [21] shows the equation for calculating CDF…”

Section: Study Of Distribution Of Workload Parameters 1) Distributmentioning

confidence: 99%

“…on the basis of resource usage it was revealed that the jobs followed a Zipf-like distribution. The PDF for zipf-like distribution is given by equation (5) [21].…”

Section: ) Distribution Of Jobs Based On Resource Usagementioning

confidence: 99%

Exploring Non-Homogeneity and Dynamicity of High Scale Cloud through Hive and Pig

Shakil

Alam

Sethi

2015

Indian Journal of Science and Technology

View full text Add to dashboard Cite

Abstract-Cloud computing deals with heterogeneity and dynamicity at all levels and therefore there is a need to manage resources in such an environment and properly allocate them. Resource planning and scheduling requires a proper understanding of arrival patterns and scheduling of resources. Study of workloads can aid in proper understanding of their associated environment. Google has released its latest version of cluster trace, trace version 2.1 in November 2014.The trace consists of cell information of about 29 days spanning across 700k jobs. This paper deals with statistical analysis of this cluster trace. Since the size of trace is very large, Hive which is a Hadoop distributed file system (HDFS) based platform for querying and analysis of Big data, has been used. Hive was accessed through its Beeswax interface. The data was imported into HDFS through HCatalog. Apart from Hive, Pig which is a scripting language and provides abstraction on top of Hadoop was used. To the best of our knowledge the analytical method adopted by us is novel and has helped in gaining several useful insights. Clustering of jobs and arrival time has been done in this paper using K-means++ clustering followed by analysis of distribution of arrival time of jobs which revealed weibull distribution while resource usage was close to zip-f like distribution and process runtimes revealed heavy tailed distribution.

show abstract

Workload Modeling for Computer Systems Performance Evaluation

Cited by 196 publications

References 620 publications

Tailbench: a benchmark suite and evaluation methodology for latency-critical applications

Tailbench: a benchmark suite and evaluation methodology for latency-critical applications

PEESOS-Cloud: A Workload-Aware Architecture for Performance Evaluation in Service-Oriented Systems

Exploring Non-Homogeneity and Dynamicity of High Scale Cloud through Hive and Pig

Contact Info

Product

Resources

About