Advanced search
2 files | 1.91 MB Add to list

BRAHMA : an intelligent framework for automated scaling of streaming and deadline-critical workflows

Ankita Atrey (UGent) , Hendrik Moens (UGent) , Gregory Van Seghbroeck (UGent) , Bruno Volckaert (UGent) and Filip De Turck (UGent)
Author
Organization
Abstract
The prevalent use of multi-component, multi-tenant models for building novel Software-as-a-Service (SaaS) applications has resulted in wide-spread research on automatic scaling of the resultant complex application workflows. In this paper, we propose a holistic solution to Automatic Workflow Scaling under the combined presence of Streaming and Deadline-critical workflows, called AWS-SD. To solve the AWS-SD problem, we propose a framework BRAHMA, that learns workflow behavior to build a knowledge-base and leverages this info to perform intelligent automated scaling decisions. We propose and evaluate different resource provisioning algorithms through CloudSim. Our results on time-varying workloads show that the proposed algorithms are effective and produce good cost-quality trade-offs while preventing deadline violations. Empirically, the proposed hybrid algorithm combining learning and monitoring, is able to restrict deadline violations to a small fraction (3-5%), while only suffering a marginal increase in average cost per component of 1-2% over our baseline naive algorithm, which provides the least costly provisioning but suffers from a large number (35-45%) of deadline violations.
Keywords
Cloud simulation, Resource provisioning, Streaming and Deadline-Critical Workflows, Deadlines, SLA

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 622.06 KB
  • brahma paper 1570285687 IEEE verified.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 1.29 MB

Citation

Please use this url to cite or link to this publication:

MLA
Atrey, Ankita et al. “BRAHMA : an Intelligent Framework for Automated Scaling of Streaming and Deadline-critical Workflows.” 2016 12TH INTERNATIONAL CONFERENCE ON NETWORK AND SERVICE MANAGEMENT AND WORKSHOPS(CNSM 2016). Oct 31-nov 04, 2016, 2016. 216–222. Print.
APA
Atrey, A., Moens, H., Van Seghbroeck, G., Volckaert, B., & De Turck, F. (2016). BRAHMA : an intelligent framework for automated scaling of streaming and deadline-critical workflows. 2016 12TH INTERNATIONAL CONFERENCE ON NETWORK AND SERVICE MANAGEMENT AND WORKSHOPS(CNSM 2016) (pp. 216–222). Presented at the 12th IEEE/IFIP/ACM International Conference on Network and Service Management (CNSM), Oct 31-nov 04, 2016.
Chicago author-date
Atrey, Ankita, Hendrik Moens, Gregory Van Seghbroeck, Bruno Volckaert, and Filip De Turck. 2016. “BRAHMA : an Intelligent Framework for Automated Scaling of Streaming and Deadline-critical Workflows.” In 2016 12TH INTERNATIONAL CONFERENCE ON NETWORK AND SERVICE MANAGEMENT AND WORKSHOPS(CNSM 2016), 216–222. Oct 31-nov 04, 2016.
Chicago author-date (all authors)
Atrey, Ankita, Hendrik Moens, Gregory Van Seghbroeck, Bruno Volckaert, and Filip De Turck. 2016. “BRAHMA : an Intelligent Framework for Automated Scaling of Streaming and Deadline-critical Workflows.” In 2016 12TH INTERNATIONAL CONFERENCE ON NETWORK AND SERVICE MANAGEMENT AND WORKSHOPS(CNSM 2016), 216–222. Oct 31-nov 04, 2016.
Vancouver
1.
Atrey A, Moens H, Van Seghbroeck G, Volckaert B, De Turck F. BRAHMA : an intelligent framework for automated scaling of streaming and deadline-critical workflows. 2016 12TH INTERNATIONAL CONFERENCE ON NETWORK AND SERVICE MANAGEMENT AND WORKSHOPS(CNSM 2016). Oct 31-nov 04, 2016; 2016. p. 216–22.
IEEE
[1]
A. Atrey, H. Moens, G. Van Seghbroeck, B. Volckaert, and F. De Turck, “BRAHMA : an intelligent framework for automated scaling of streaming and deadline-critical workflows,” in 2016 12TH INTERNATIONAL CONFERENCE ON NETWORK AND SERVICE MANAGEMENT AND WORKSHOPS(CNSM 2016), Ecole Technologie Superiere, Montreal, CANADA, 2016, pp. 216–222.
@inproceedings{8557923,
  abstract     = {{The prevalent use of multi-component, multi-tenant models for building novel Software-as-a-Service (SaaS) applications has resulted in wide-spread research on automatic scaling of the resultant complex application workflows. In this paper, we propose a holistic solution to Automatic Workflow Scaling under the combined presence of Streaming and Deadline-critical workflows, called AWS-SD. To solve the AWS-SD problem, we propose a framework BRAHMA, that learns workflow behavior to build a knowledge-base and leverages this info to perform intelligent automated scaling decisions. We propose and evaluate different resource provisioning algorithms through CloudSim. Our results on time-varying workloads show that the proposed algorithms are effective and produce good cost-quality trade-offs while preventing deadline violations. Empirically, the proposed hybrid algorithm combining learning and monitoring, is able to restrict deadline violations to a small fraction (3-5%), while only suffering a marginal increase in average cost per component of 1-2% over our baseline naive algorithm, which provides the least costly provisioning but suffers from a large number (35-45%) of deadline violations.}},
  author       = {{Atrey, Ankita and Moens, Hendrik and Van Seghbroeck, Gregory and Volckaert, Bruno and De Turck, Filip}},
  booktitle    = {{2016 12TH INTERNATIONAL CONFERENCE ON NETWORK AND SERVICE MANAGEMENT AND WORKSHOPS(CNSM 2016)}},
  isbn         = {{978-3-901882-85-2}},
  issn         = {{2165-9605}},
  keywords     = {{Cloud simulation,Resource provisioning,Streaming and Deadline-Critical Workflows,Deadlines,SLA}},
  language     = {{eng}},
  location     = {{Ecole Technologie Superiere, Montreal, CANADA}},
  pages        = {{216--222}},
  title        = {{BRAHMA : an intelligent framework for automated scaling of streaming and deadline-critical workflows}},
  year         = {{2016}},
}

Web of Science
Times cited: