13

Click here to load reader

Software-Defined Inter-Cloud Composition of Big Services

Embed Size (px)

Citation preview

Page 1: Software-Defined Inter-Cloud Composition of Big Services

Pradeeban Kathiravelu(INESC-ID/IST ULisboa, Portugal and UCLouvain, Belgium)

Supervised By: Prof. Lu s Veiga (INESC-ID/IST ULisboa) ıı Prof. Marco Canini (KAUST)

Prof. Peter Van Roy (UCLouvain)

Software-Defined Inter-Cloud Composition of Big Services

EMJD-DC Spring Event, BrusselsMay 30, 2017

Page 2: Software-Defined Inter-Cloud Composition of Big Services

2/12

Introduction

● Big services with complex workloads.

– Geographically distributed big data.

– Resource availabilities at remote locations; for example,

● Distributed clouds.

● Volunteer computing.

● Edge computing.

● Service Level Agreements (SLAs) in multi-tenant clouds.

Page 3: Software-Defined Inter-Cloud Composition of Big Services

3/12

Challenges

● Big services in multi-tenant environments.

– Differentiated QoS in cloud networks.

● Can we discriminate the services with redundancy in data and execution paths?

Page 4: Software-Defined Inter-Cloud Composition of Big Services

4/12

Motivation● Componentize Big Services in the Internet.

● Efficient network-aware composition and execution of big services.

Page 5: Software-Defined Inter-Cloud Composition of Big Services

5/12

Contributions● An inter-cloud framework to componentize and compose

big services.

– Execute them as a network-aware distributed

service composition.

– Modelling, Scalability, and orchestration.

● Use the current best-fit execution path.

– Web services and microservices as the

building blocks of the big services.

Page 6: Software-Defined Inter-Cloud Composition of Big Services

6/12

Approach● Network-level guarantees based on

application/service-level inputs.

– Synergy of services with network.

Page 7: Software-Defined Inter-Cloud Composition of Big Services

7/12

Mayan (Componentizing Big Services)

● A scalable resilient framework for inter-domain big service execution.

Page 8: Software-Defined Inter-Cloud Composition of Big Services

8/12

Prototype Assessments● Increased QoS and Speedup.

– Performance growth =

f(problem size, workflow as services).

● Network-aware scalability and distribution.

● Minimize communication and coordination overheads.

Page 9: Software-Defined Inter-Cloud Composition of Big Services

9/12

Conclusion● Summary

– Synergy of network and service level properties in big service execution.

– Componentizing big services and execute as service compositions.

– Scalability and resilience for multi-tenant distributed clouds.

– A federated controller deployment to orchestrate inter-cloud networks.

● Future Work

– Leverage cloud providers for differentiated connectivity.

– Big Services to the Edge.

Page 10: Software-Defined Inter-Cloud Composition of Big Services

10/12

Publications (Since 2016 Spring Event)

Book Chapter

● Kathiravelu, P. & Veiga, L. (2017). SDN helps other Vs in Big Data. Chapter of Big Data and Software Defined Networks. Sep. 2017. 21 pages. (Camera-ready copy submitted).

Page 11: Software-Defined Inter-Cloud Composition of Big Services

11/12

Conference Proceedings● Kathiravelu, P. & Veiga, L. (2017). SD-CPS: Taming the Challenges of Cyber-Physical Systems with a Software-

Defined Approach. In The 4th International Conference on Software Defined Systems (SDS-2017). May 2017. 8 pages. To Appear.

● Kathiravelu, P. & Veiga, L. (2017). SDN Middlebox Architecture for Resilient Transfers. In 15th IFIP/IEEE International Symposium on Integrated Network Management (IM 2017), May 2017. 4 pages. To Appear.

● Kathiravelu, P. & Veiga, L. (2016). Software-Defined Simulations for Continuous Development of Cloud and Data Center Networks. In 24 th International Conference on Cooperative Information Systems (CoopIS 2016). On the Move to Meaningful Internet Systems: OTM 2016 Conferences, pp. 3 – 23. Springer International Publishing, Oct. 2016.

● Kathiravelu, P., & Veiga, L. (2016). Selective Redundancy in Network-as-a-Service: Differentiated QoS in Multi-tenant Clouds. In OTM Confederated International Conferences. On the Move to Meaningful Internet Systems (pp. 87-97). EI2N 2016. Springer, Cham.

● Kathiravelu, P. & Sharma, A. (2016). A Dynamic Data Warehousing Platform for Creating and Accessing Biomedical Data Lakes. In 2nd International Workshop on Data Management and Analytics for Medicine and Healthcare (DMAH'16), co-located with 42 nd International Conference on Very Large Data Bases. Sep. 2016. LNCS. pp. 101 – 120.

● Caixinha, D., Kathiravelu, P. & Veiga, L. (2016). ViTeNA: An SDN-Based Virtual Network Embedding Algorithm for Multi-Tenant Data Centers. In 15 th IEEE International Symposium on Network Computing and Applications (NCA 2016), Oct. 2016. pp. 140 – 147.

Page 12: Software-Defined Inter-Cloud Composition of Big Services

12/12

Conference Proceedings● Kathiravelu, P. & Veiga, L. (2017). SD-CPS: Taming the Challenges of Cyber-Physical Systems with a Software-

Defined Approach. In The 4th International Conference on Software Defined Systems (SDS-2017). May 2017. 8 pages. To Appear.

● Kathiravelu, P. & Veiga, L. (2017). SDN Middlebox Architecture for Resilient Transfers. In 15th IFIP/IEEE International Symposium on Integrated Network Management (IM 2017), May 2017. 4 pages. To Appear.

● Kathiravelu, P. & Veiga, L. (2016). Software-Defined Simulations for Continuous Development of Cloud and Data Center Networks. In 24 th International Conference on Cooperative Information Systems (CoopIS 2016). On the Move to Meaningful Internet Systems: OTM 2016 Conferences, pp. 3 – 23. Springer International Publishing, Oct. 2016.

● Kathiravelu, P., & Veiga, L. (2016). Selective Redundancy in Network-as-a-Service: Differentiated QoS in Multi-tenant Clouds. In OTM Confederated International Conferences. On the Move to Meaningful Internet Systems (pp. 87-97). EI2N 2016. Springer, Cham.

● Kathiravelu, P. & Sharma, A. (2016). A Dynamic Data Warehousing Platform for Creating and Accessing Biomedical Data Lakes. In 2nd International Workshop on Data Management and Analytics for Medicine and Healthcare (DMAH'16), co-located with 42 nd International Conference on Very Large Data Bases. Sep. 2016. LNCS. pp. 101 – 120.

● Caixinha, D., Kathiravelu, P. & Veiga, L. (2016). ViTeNA: An SDN-Based Virtual Network Embedding Algorithm for Multi-Tenant Data Centers. In 15 th IEEE International Symposium on Network Computing and Applications (NCA 2016), Oct. 2016. pp. 140 – 147.

Thank you.

Questions?

Page 13: Software-Defined Inter-Cloud Composition of Big Services

13/12

Announcement!

● Please join the Erasmus Mundus Association– em-a.eu