Task-based programming in COMPSs to converge from HPC to big data

Abstract

Task-based programming has proven to be a suitable model for high-performance computing (HPC) applications. Different implementations have been good demonstrators of this fact and have promoted the acceptance of task-based programming in the OpenMP standard. Furthermore, in recent years, Apache Spark has gained wide popularity in business and research environments as a programming model for addressing emerging big data problems. COMP Superscalar (COMPSs) is a task-based environment that tackles distributed computing (including Clouds) and is a good alternative for a task-based programming model for big data applications. This article describes why we consider that task-based programming models are a good approach for big data applications. The article includes a comparison of Spark and COMPSs in terms of architecture, programming model, and performance. It focuses on the differences that both frameworks have in structural terms, on their programmability interface, and in terms of their efficiency by means of three widely known benchmarking kernels: Wordcount, Kmeans, and Terasort. These kernels enable the evaluation of the more important functionalities of both programming models and analyze different work flows and conditions. The main results achieved from this comparison are (1) COMPSs is able to extract the inherent parallelism from the user code with minimal coding effort as opposed to Spark, which requires the existing algorithms to be adapted and rewritten by explicitly using their predefined functions, (2) it is an improvement in terms of performance when compared with Spark, and (3) COMPSs has shown to scale better than Spark in most cases. Finally, we discuss the advantages and disadvantages of both frameworks, highlighting the differences that make them unique, thereby helping to choose the right framework for each particular objective.

Keywords

Programming models distributed computing framework comparison Big Data programming

Get full access to this article

View all access options for this article.

References

Amazon Enterprise Cloud Computing (2016) Available at: http://aws.amazon.com/enterprise/ (accessed 27 November 2016).

Amazon Simple Storage Service (S3) (2016) Available at: https://aws.amazon.com/s3/ (accessed 20 December 2016).

Apache Hadoop (2016) Available at: http://hadoop.apache.org/ (accessed 15 November 2016).

Apache Spark – Lightning-fast cluster computing (2016) Available at: http://spark.apache.org/ (accessed 15 November 2016).

Badia

Conejero

Diaz

. (2015) {COMP} superscalar, an interoperable programming framework. SoftwareX, 34: 32–36.

COMP Superscalar (2016) Available at: http://compss.bsc.es/ (accessed 17 November 2016).

Da Silva Morais

(2015) Survey on frameworks for distributed computing: Hadoop, Spark and Storm. In: De Sousa

Oliveira

(eds) Proceedings of the 10th Doctoral Symposium in Informatics Engineering, Porto (Portugal), 29–30 January, 2015. pp. 95–105, 2015. Porto, Portugal: Faculdade de Engenharia da Universidade do Porto.

Decorators for Functions and Methods (2016) Available at: https://www.python.org/dev/peps/pep-0318/ (accessed 17 November 2016).

Ekanayake

Kamburugamuve

Wickramasinghe

. (2016) Java thread and process performance for parallel machine learning on multicore HPC clusters. In: Proceedings of the 2016 IEEE International Conference on Big Data Washington, DC, USA, 5–8 December 2016.

10.

ehigss Spark Terasort (2016) Available at: https://github.com/ehiggs/spark-terasort (accessed 25 November 2016).

11.

Examples Apache Spark (2016) Available at: http://spark.apache.org/examples.html (accessed 22 November 2016).

12.

Extrae (2016) Available at: https://tools.bsc.es/extrae (accessed 19 December 2016).

13.

Fox

Qiu

Jha

Ekanayake

(2015). Big data, simulations and HPC convergence. In: Rabl

Nambiar

Baru

. (eds) Big Data Benchmarking – 6th International Workshop, WBDB 2015, Toronto, ON, Canada, June 16–17, 2015 and 7th International Workshop, WBDB 2015, New Delhi, India, December 14–15, 2015, Revised Selected Papers, pp. 3–17, 2015. Cham (ZG), Switzerland: Springer International Publishing.

14.

Gopalani

Arora

(2015) Comparing apache spark and map reduce with performance analysis using K-Means. International Journal of Computer Applications 113(1): 8–11.

15.

Google Cloud Platform – Cloud Computing, Hosting Services & APIs (2016) Available at: http://cloud.google.com/ (accessed 4 December 2016).

16.

Hindman

Konwinski

Zaharia

. (2011) Mesos: a platform for fine-grained resource sharing in the data center. In: Proceedings of the 8th USENIX Conference on Networked Systems Design and Implementation, NSDI’11, Boston, Massachusetts, USA, 30 March–1 April 2011, pp. 295–308. Berkeley, CA, USA: USENIX Association.

17.

JNI APIs and Developer Guides (2016) Available at: https://docs.oracle.com/javase/8/docs/technotes/guides/jni/ (accessed 19 December 2016).

18.

Liang

(1999) Java Native Interface: Programmer’s Guide and Reference, 1st ed. Boston: Addison-Wesley Longman Publishing Co., Inc.

19.

MareNostrum BSC-CNS (2016) Available at: https://www.bsc.es/innovation-and-services/supercomputers-and-facilities/marenostrum (accessed 22 November 2016).

20.

Marti

Gasull

Queralt

. (2013) Towards DaaS 2.0: enriching data models. In: Proceedings – 2013 IEEE 9th World Congress on Services, SERVICES 2013, Santa Clara, California, 28 June–3 July 2013, June 2013, pp. 349–355. IEEE.

21.

Meng

Bradley

Yavuz

. (2016) Mllib: machine learning in apache spark. Journal of Machine Learning Research 17(1): 1235–1241.

22.

Moreno-Vozmediano

Montero

Llorente

. (2012) IaaS cloud architecture: from virtualized datacenters to federated cloud infrastructures. Computer 45(12): 65–72.

23.

OCCI – Open Cloud Computing Interface (2016) Available at: http://occi-wg.org/ (accessed 27 November 2016).

24.

OpenNebula – Flexible Enterprise Cloud Made Simple (2016) Available at: http://opennebula.org/ (accessed 15 November 2016).

25.

OpenStack – Open Source Cloud Computing Software (2016) Available at: http://openstack.org/ (accessed 27 November 2016).

26.

Paraver: a flexible performance analysis tool. Available at: https://tools.bsc.es/paraver, (accessed 19 December 2016).

27.

Pan

(2016) The performance comparison of hadoop and spark. Master’s Thesis, St. Cloud State University, USA. Culminating Projects in Computer Science and Information Technology. Paper 7.

28.

Pillet

Labarta

Cortes

. (1995) Paraver: a tool to visualize and analyze parallel code. Transputer and Occam Developments. 17–32, April 1995. Available at: http://www.bsc.es/ (accessed 17–32 April 2012).

29.

Reed

Dongarra

(2015) Exascale computing and big data. ACM Communications 58(7): 56–68.

30.

Reyes-Ortiz

Oneto

Anguita

(2015) Big Data analytics in the cloud: spark on Hadoop vs MPI/OpenMP on Beowulf. Procedia Computer Science 53(Complete): 121–130.

31.

Shi

Qiu

Minhas

Jiao

. (2015) Clash of the titans: mapreduce vs. spark for large scale data analytics. Proceedings of the VLDB Endowment 8(13): 2110–2121.

32.

Shvachko

Kuang

Radia

. (2010) The hadoop distributed file system. In: Proceedings of the 2010 IEEE 26th Symposium on Mass Storage Systems and Technologies (MSST), MSST’10, Washington, DC, USA, May 2010, pp. 1–10. Washington: IEEE Computer Society.

33.

Tejedor

Becerra

Alomar

. (2015) PyCOMPSs: parallel computational workflows in Python. International Journal of High Performance Computing Applications 31(1): 66–82.

34.

UCI Machine Learning Repository: HIGGS Data Set (2016) Available at: http://archive.ics.uci.edu/ml/datasets/HIGGS (accessed 4 December 2016).

35.

Vavilapalli

Murthy

Douglas

. (2013) Apache hadoop yarn: yet another resource negotiator. In: Proceedings of the 4th Annual Symposium on Cloud Computing, SOCC’13, New York, NY, USA, October 2013, pp. 1–5, 16. New York, NY: ACM.

36.

Zaharia

Chowdhury

Franklin

. (2010) Spark: cluster computing with working sets. In: Proceedings of the 2nd USENIX Conference on Hot Topics in Cloud Computing, HotCloud’10, Berkeley, CA, USA, June 2010. Berkeley: USENIX Association.