DAReSch: deadline-aware request scheduling for cloud storage services

Abstract

With the emergence of cloud computing and big data, many companies are increasingly relying on the cloud to store and retrieve tremendous amounts of data to leverage the scalability and performance offered by cloud storage services. As a result, data retrieval time has become of a paramount importance for cloud users especially. However, current data management systems are still not optimized to reduce such time. In this paper, we present a deadline-aware data request scheduling scheme, called DAReSch, that aims at scheduling data requests in order to minimize data transfer times and to meet the deadlines specified by the users. We show through real experiments using OpenStack storage system (i.e., Swift) that, compared with the traditional Swift Client system, DAReSch significantly increases the percentage of requests meeting their deadlines, reduces data transfer time, and maximizes bandwidth usage. Furthermore, we also study the impact of the request deadlines on the studied performance metrics. Our extensive experiments show that, when deadlines are stringent, DAReSh allows 60% of the requests to meet their deadline requirements compared with only 10% for the existing solution. When the deadline is less stringent, 90% of the requests can meet their deadline requirements with DAReSh compared with 40% for the existing solution.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10

References

  1. 1.

    Amazon s3. https://aws.amazon.com/fr/s3/

  2. 2.

    Dropbox. https://www.dropbox.com/

  3. 3.

    Google drive. https://drive.google.com/

  4. 4.

    Cloud storage market. https://www.marketsandmarkets.com/-PressReleases/cloud-storage.asp

  5. 5.

    Linden G (2006) Make data useful. Stanford CS345 Talk

  6. 6.

    Schurman E, Brutlag J (2009) The user and business impact of server delays, additional bytes, and http chunking in web search. O’Reilly Web Performance & Operations Converence

  7. 7.

    Ghemawat S, Gobioff H, Leung ST (2003) The google file system. ACM SOSP

  8. 8.

    Tlili G, Zhani MF, Elbiaze H (2018) On providing deadline-aware Cloud storage services. In: IEEE conference on innovation in clouds, internet and networks (ICIN 2018), Paris

  9. 9.

    Dieye M, Zhani MF, Elbiaze H (2017) On achieving high data availability in heterogeneous cloud storage systems. In: 2017 IFIP/IEEE Symposium on Integrated Network and Service Management (IM)

  10. 10.

    Joshi G, Soljanin E, Wornell G (2015) Efficient replication of queued tasks for latency reduction in cloud systems. In: IEEE annual Allerton conference on communication, control, and computing

  11. 11.

    Joshi G, Soljanin E, Wornell G (2015) Efficient replication of queued tasks for latency reduction in cloud systems. In: IEEE annual Allerton conference on communication, control, and computing

  12. 12.

    Wang D, Joshi G, Wornell G (2015) Using straggler replication to reduce latency in large-scale parallel computing. ACM SIGMETRICS Performance Evaluation Review 43(3):7–11

    Article  Google Scholar 

  13. 13.

    Liu W, Tieman B, Kettimuthu R, Foster I (2010) A data transfer framework for large-scale science experiments. In: ACM international symposium on high performance distributed computing

  14. 14.

    Hou B, Chen F, Ou Z, Wang R, Mesnier M (2017) Understanding i/o performance behaviors of cloud storage from a client’s perspective. ACM Trans Storage (TOS) 13(2):16

    Google Scholar 

  15. 15.

    Liu G, Shen H, Yu L (2016) Towards deadline guaranteed cloud storage services. In: IEEE international conference on cloud computing (CLOUD)

  16. 16.

    Kosar T, Arslan E, Ross B, Zhang B (2013) Storkcloud: data transfer scheduling and optimization as a service. In: ACM workshop on scientific cloud computing

  17. 17.

    Janet J, Balakrishnan S, Murali E (2016) Improved data transfer scheduling and optimization as a service in cloud. In: IEEE international conference on information communication and embedded systems (ICICES), pp 1–3

  18. 18.

    Zhani MF, Elbiaze H, Kamoun F (2008) Analysis of prediction performance of training-based models using real network traffic. In: International symposium on performance evaluation of computer and telecommunication systems, pp 472–479

  19. 19.

    Zhani MF, Elbiaze H, Kamoun F (2009) Analysis and prediction of real network traffic. JNW 4 (9):855–865

    Article  Google Scholar 

  20. 20.

    Openstack swift object storage service. http://swift.openstack.org/

  21. 21.

    Rackspace. https://www.rackspace.com/

  22. 22.

    ebay. www.ebay.com/

  23. 23.

    Instagram. https://www.instagram.com/

  24. 24.

    Object size limit. https://aws.amazon.com/blogs/aws/amazon-s3-object-size-limit/

  25. 25.

    Python-swiftclient. https://docs.openstack.org/developer/python-swiftclient/introduction.html/

  26. 26.

    Yahoo! webscope dataset: statistical information regarding files and access pattern to files in one of yahoo clusters. http://webscope.sandbox.yahoo.com/catalog.php?datatype=s/, Accessed 19 Apr 2017

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Mohamed Faten Zhani.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Tlili, G., Yahyaoui, H., Zhani, M.F. et al. DAReSch: deadline-aware request scheduling for cloud storage services. Ann. Telecommun. 74, 545–557 (2019). https://doi.org/10.1007/s12243-019-00728-4

Download citation

Keywords

  • Cloud storage systems
  • OpenStack Swift
  • Request scheduling
  • Deadline-aware data retrieval