Advertisement

Non-linear analysis of bursty workloads using dual metrics for better cloud resource management

  • Mahesh BalajiEmail author
  • Ch. Aswani Kumar
  • G. Subrahmanya V. R. K. Rao
Original Research
  • 20 Downloads

Abstract

The assumption that enterprise workloads are steady-state could make their resource provisioning ineffective. The current study aims to address this challenge by performing a non-linear analysis on a set of synthetic bursty workloads. The research involves building resource-provisioning models using non-linear metrics, namely hurst exponent and sample entropy. Performance of the proposed approach was compared with baseline reactive approach and the index of dispersion approach using the NASA dataset. The proposed approach had a sensitivity of 70% and specificity of 90%. The reactive approach had a sensitivity and specificity of 55% and 84%, respectively while the index of dispersion had a sensitivity of 61% and specificity of 82%. The current study also displayed a 71% reduction in error-count compared to the baseline reactive approach.

Keywords

Bursty workload Cloud computing Non-linear workload Resource management 

Notes

References

  1. Alam M, Shakil KA, Sethi S (2015) Analysis and clustering of workload in google cluster trace based on resource usage. 2016 IEEE Intl conference on computational science and engineering (CSE) and IEEE Intl conference on embedded and ubiquitous computing (EUC) and 15th Intl symposium on distributed computing and applications for business engineering (DCABES), pp 740–747.  https://doi.org/10.1109/CSE-EUC-DCABES.2016.271
  2. Ali-eldin A, Tordsson J, Elmroth E, Kihl M (2013) Workload classification for efficient auto-scaling of cloud resources, pp 1–36. https://pdfs.semanticscholar.org/c808/e997f6c32237ce0cb5a0f0eacf72c46e4611.pdf. Accessed 28 July 2017
  3. Ali-Eldin A, Seleznjev O, Sjostedt-de Luna S, Tordsson J, Elmroth E (2014) Measuring cloud workload burstiness. In 2014 IEEE/ACM 7th international conference on utility and cloud computing (vol 2014, pp 566–572). IEEE.  https://doi.org/10.1109/UCC.2014.87
  4. Amazon Web Services, Inc (2016) T2 instances. http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/t2-instances.html
  5. Arlitt M, Jin T (2000) A workload characterization study of the 1998 World Cup Web site. IEEE Netw 14(3):30–37.  https://doi.org/10.1109/65.844498 CrossRefGoogle Scholar
  6. Arlitt MF, Williamson CL (1997) Internet Web servers: workload characterization and performance implications. IEEE/ACM Trans Netw 5(5):631–645.  https://doi.org/10.1109/90.649565 CrossRefGoogle Scholar
  7. Armbrust M, Stoica I, Zaharia M, Fox A, Griffith R, Joseph AD et al (2010) A view of cloud computing. Commun ACM 53(4):50.  https://doi.org/10.1145/1721654.1721672 CrossRefGoogle Scholar
  8. Balaji M, Kumar CA, Rao GSVRK (2016) Predictive Cloud resource management framework for enterprise workloads. J King Saud Univ Comput Inf Sci.  https://doi.org/10.1016/j.jksuci.2016.10.005 Google Scholar
  9. Bodik P, Fox A, Franklin MJ, Jordan MI, Patterson DA (2010) Characterizing, modeling, and generating workload spikes for stateful services. In: Proceedings of the 1st ACM symposium on Cloud computing—SoCC’10 (vol d, p 241). ACM Press, New York.  https://doi.org/10.1145/1807128.1807166
  10. Bougadis J, Adamowski K, Diduch R (2005) Short-term municipal water demand forecasting. Hydrol Process 19(1):137–148.  https://doi.org/10.1002/hyp.5763 CrossRefGoogle Scholar
  11. Caniff A, Lu L, Mi N, Cherkasova L, Smirni E (2010) Fastrack for taming burstiness and saving power in multi-tiered systems. In: 2010 22nd International Teletraffic Congress (lTC 22) (p 1–8). IEEE.  https://doi.org/10.1109/ITC.2010.5608741
  12. Casale G, Mi N, Cherkasova L, Smirni E (2008) How to parameterize models with bursty workloads. ACM SIGMETRICS Perform Eval Rev 36(2):38.  https://doi.org/10.1145/1453175.1453182 CrossRefGoogle Scholar
  13. Cecchet E (2009) RUBiS: Rice University Bidding System. http://rubis.ow2.org/index.html. Accessed 8 Dec 2017
  14. Chen Y, Ganapathi A (2010) Towards understanding cloud performance tradeoffs using statistical workload analysis and replay. University of California at. http://www.eecs.berkeley.edu/Pubs/TechRpts/2010/EECS-2010-81.pdf. Accessed 3 Sept 2017
  15. Clegg RG (2006) A practical guide to measuring the hurst parameter. Int J Simul Syst Sci Technol 7(2):3–14. http://arxiv.org/abs/math/0610756. Accessed 28 July 2017
  16. Ding S, Li Y, Wu D, Zhang Y, Yang S (2018) Time-aware cloud service recommendation using similarity-enhanced collaborative filtering and ARIMA model. Decis Support Syst 107:103–115.  https://doi.org/10.1016/j.dss.2017.12.012 CrossRefGoogle Scholar
  17. Eldin AA, Rezaie A, Mehta A, Razroev S, De Luna SS, Seleznjev O et al (2014) How will your workload look like in 6 years? Analyzing Wikimedia’s Workload. In: 2014 IEEE international conference on cloud engineering (pp 349–354). IEEE.  https://doi.org/10.1109/IC2E.2014.50
  18. Exponent TH, Series T (2015) The hurst exponent: predictability of time series. http://analytics-magazine.org/the-hurst-exponent-predictability-of-time-series. Accessed 12 Oct 2015
  19. Forman G (2004) A pitfall and solution in multi-class feature selection for text classification. In: Proceedings of the twenty-first international conference on machine learning (p 38). ACM Press, New York.  https://doi.org/10.1145/1015330.1015356
  20. George AD, Squillace R, Nottingham C, Pratt T (2017) How to configure auto scaling for a Cloud Service in the classic portal. https://docs.microsoft.com/en-us/azure/cloud-services/cloud-services-how-to-scale. Accessed 8 Dec 2017
  21. Gusella R (1991) Characterizing the variability of arrival processes with indexes of dispersion. IEEE J Sel Areas Commun 9(2):203–211.  https://doi.org/10.1109/49.68448 CrossRefGoogle Scholar
  22. Hua G, Zhang M, Liu Y, Ma S, Ru L (2010) Hierarchical feature selection for ranking. In: Proceedings of the 19th international conference on world wide web—WWW’10 (p 1113). ACM Press, New York.  https://doi.org/10.1145/1772690.1772830
  23. Lassnig M, Fahringer T, Garonne V, Molfetas A, Branco M (2010) Identification, modelling and prediction of non-periodic bursts in workloads. In: 2010 10th IEEE/ACM international conference on cluster, cloud and grid computing (pp 485–494). IEEE.  https://doi.org/10.1109/CCGRID.2010.118
  24. Leland WE, Taqqu MS, Willinger W, Wilson DV (1993) On the self-similar nature of Ethernet traffic. ACM SIGCOMM Comput Commun Rev 23(4):183–193.  https://doi.org/10.1145/167954.166255 CrossRefGoogle Scholar
  25. Lorido-Botran T, Huerta S, Tomás L, Tordsson J, Sanz B (2017) An unsupervised approach to online noisy-neighbor detection in cloud data centers. Expert Syst Appl 89:188–204.  https://doi.org/10.1016/j.eswa.2017.07.038 CrossRefGoogle Scholar
  26. Lyman Ott R, Longnecker M (2013) An introduction to statistical methods and data analysis, 7th edn. CENGAGE Learning, BostonGoogle Scholar
  27. Manikandan S (2011) Measures of dispersion. J Pharmacol Pharmacother 2(4):315.  https://doi.org/10.4103/0976-500X.85931 CrossRefGoogle Scholar
  28. Mao M, Humphrey M (2012) A performance study on the VM startup time in the cloud. In: Proceedings—2012 IEEE 5th international conference on cloud computing, CLOUD 2012, (Cloud), pp 423–430.  https://doi.org/10.1109/CLOUD.2012.103
  29. Mao M, Li J, Humphrey M (2010) Cloud auto-scaling with deadline and budget constraints. In: 2010 11th IEEE/ACM international conference on grid computing (pp 41–48). IEEE.  https://doi.org/10.1109/GRID.2010.5697966
  30. Marshall P, Keahey K, Freeman T (2010) Elastic site: using clouds to elastically extend site resources. In: 2010 10th IEEE/ACM international conference on cluster, cloud and grid computing (pp 43–52). IEEE.  https://doi.org/10.1109/CCGRID.2010.80
  31. Mehta A, Durango J, Tordsson J, Elmroth E (2015) Online spike detection in cloud workloads. In: 2015 IEEE international conference on cloud engineering (pp 446–451). IEEE.  https://doi.org/10.1109/IC2E.2015.50
  32. Mell PM, Grance T (2011) The NIST definition of cloud computing. In: Special publication (NIST SP)—800—145 (vol 28). The National Institute of Standards and Technology (NIST). http://csrc.nist.gov/publications/PubsSPs.html#800-145. Accessed 26 July 2017
  33. Mi N, Casale G, Cherkasova L, Smirni E (2008) Burstiness in multi-tier applications: symptoms, causes, and new models. Lecture Notes Comput Sci (Including Subser Lecture Notes Artif Intell Lecture Notes Bioinform) 5346:265–286.  https://doi.org/10.1007/978-3-540-89856-6_14 Google Scholar
  34. Mishra AK, Hellerstein JL, Cirne W, Das CR (2010) Towards characterizing cloud backend workloads. ACM SIGMETRICS Perform Eval Rev 37(4):34.  https://doi.org/10.1145/1773394.1773400 CrossRefGoogle Scholar
  35. Montanari A, Taqqu MS, Teverovsky V (1999) Estimating long-range dependence in the presence of periodicity: an empirical study. Math Comput Model 29(10–12):217–228.  https://doi.org/10.1016/S0895-7177(99)00104-1 CrossRefzbMATHGoogle Scholar
  36. Moreno IS, Garraghan P, Townend P, Xu J (2013) An approach for characterizing workloads in google cloud to derive realistic resource utilization models. In: Proceedings—2013 IEEE 7th international symposium on service-oriented system engineering, SOSE 2013 (vol 60, pp 49–60). IEEE.  https://doi.org/10.1109/SOSE.2013.24
  37. Mulia W, Acken J, Fritz D, Sehgal N, Sohoni S, Stanberry C (2013) Cloud workload characterization. IETE Tech Rev 30(5):382.  https://doi.org/10.4103/0256-4602.123121 CrossRefGoogle Scholar
  38. Naghibi SA (2016) Information gain ratio. https://en.wikipedia.org/wiki/Information_gain_ratio. Accessed 8 Dec 2017
  39. Otto L, Hurst T (2012) Hurst exponent. https://en.wikipedia.org/wiki/Hurst_exponent. Accessed 8 Dec 2017
  40. Palshikar GK (2009) Simple algorithms for peak detection in time-series simple algorithms for peak detection in time-series. Technical report, TRDDC, (August). https://www.researchgate.net/publication/228853276%7B_%7DSimple%7B_%7DAlgorithms%7B_%7Dfor%7B_%7DPeak%7B_%7DDetection%7B_%7Din%7B_%7DTime-Series. Accessed 11 Aug 2017
  41. Papadopoulos AV, Ali-Eldin A, Arzen K-E, Tordsson J, Elmroth E (2016) PEAS. ACM Trans Model Perform Eval Comput Syst 1(4):1–31.  https://doi.org/10.1145/2930659 CrossRefGoogle Scholar
  42. Perez-Palacin D, Mirandola R, Merseguer J (2017) Accurate modeling and efficient QoS analysis of scalable adaptive systems under bursty workload. J Syst Softw 130:24–41.  https://doi.org/10.1016/j.jss.2017.05.022 CrossRefGoogle Scholar
  43. Saripalli P, Kiran GVR, Shankar RR, Narware H, Bindal N (2011) Load Prediction and Hot Spot Detection Models for Autonomic Cloud Computing. In: 2011 fourth IEEE international conference on utility and cloud computing (pp 397–402). IEEE.  https://doi.org/10.1109/UCC.2011.66
  44. Singh R, Sharma U, Cecchet E, Shenoy P (2010) Autonomic mix-aware provisioning for non-stationary data center workloads. In: Proceeding of the 7th international conference on autonomic computing—ICAC’10 (p 21). ACM Press, New York.  https://doi.org/10.1145/1809049.1809053
  45. Tai J, Zhang J, Li J, Meleis W, Mi N (2011) ArA: adaptive resource allocation for cloud computing environments under bursty workloads. In: 30th IEEE international performance computing and communications conference (pp 1–8). IEEE.  https://doi.org/10.1109/PCCC.2011.6108060
  46. Tulloch M (2013) Introducing Windows Azure for IT Professionals. Microsoft Press, SeattleGoogle Scholar
  47. Varia J (2010) Architecting for the Cloud: best practices. Amazon Web Services. http://media.amazonwebservices.com/AWS_Cloud_Best_Practices.pdf. Accessed 8 July 2017
  48. Vlachos M, Meek C, Vagena Z, Gunopulos D (2004) Identifying similarities, periodicities and bursts for online search queries. In: 2004 ACM SIGMOD int. conf. on management of data. ACM Press, New York, pp 131–142.  https://doi.org/10.1145/1007568.1007586 CrossRefGoogle Scholar
  49. Wang C, Talwar V, Schwan K, Ranganathan P (2010) Online detection of utility cloud anomalies using metric distributions. In: Proceedings of the 2010 IEEE/IFIP network operations and management symposium, NOMS 2010 (pp 96–103). IEEE.  https://doi.org/10.1109/NOMS.2010.5488443
  50. Wang Q, Kanemasa Y, Kawaba M, Pu C (2012) When average is not average. In: Proceedings of the 9th international conference on autonomic computing—ICAC’12 (p 33). ACM Press, New York.  https://doi.org/10.1145/2371536.2371544
  51. Wang J, Dou W, Gao C, Wei J (2018) JSTrace: fast reproducing web application errors. J Syst Softw 137:448–462.  https://doi.org/10.1016/j.jss.2017.06.038 CrossRefGoogle Scholar
  52. Wildfire A (2017) Going viral. In: EBR—European biopharmaceutical review. ACM Press, New York, pp 16–20.  https://doi.org/10.1145/2068816.2068867 Google Scholar
  53. Youssef AA, Krishnamurthy D (2017) Burstiness-aware service level planning for enterprise application clouds. J Cloud Comput 6(1):17.  https://doi.org/10.1186/s13677-017-0087-y CrossRefGoogle Scholar
  54. Zhang S, Qian Z, Luo Z, Wu J, Lu S (2016) Burstiness-aware resource reservation for server consolidation in computing clouds. IEEE Trans Parallel Distrib Syst 27(4):964–977.  https://doi.org/10.1109/TPDS.2015.2425403 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Global Technology OfficeCognizant Technology SolutionsChennaiIndia
  2. 2.School of Information Technology and EngineeringVIT UniversityVelloreIndia

Personalised recommendations