Workshop Program

[18/08/17]

14:00 - 14:05 Welcome, E. Casalicchio

14:05 - 14:30 Towards Vertically Scalable Spark Applications. G. Quattrocchi, L. Baresi.

14:30 - 15:00 A Resource Allocation Framework with Qualitative and Quantitative SLA Classes. C. Cerin, T. Menouer, W. Saad and X. Shi

15:00 - 15:30 The Impact of the Storage Tier: A Baseline Performance Analysis of containerized DBMS. D. Seybold, C. B. Hauser, G. Eisenhart, S. Volpert and J. Domaschka

15-30-16:00 Coffee break

16:00 - 17:00. Invited Talk - Self-adaptation for Streaming Analytics at the Edge, Valeria Cardellini, University of Rome Tor Vergata, http://www.ce.uniroma2.it/~valeria/

17:00 - 17:30. Automated Multi-Swarm Networking with Open Baton NFV MANO Framework. J-S. Shin, M. Santos de Brito, T. Magedanz and J. Kim

Invited talk - Self-adaptation for Streaming Analytics at the Edge

Valeria Cardellini, University of Rome Tor Vergata

Abstract: Data Stream Processing (DSP) systems are commonly used to process big data streams from sensors and devices and there is a need to push streaming analytics capabilities to the edges of the network in order to cut down the latency. This scenario requires to devise effective solutions to manage and self-adapt at run-time the execution of DSP applications in the presence of unforeseeable variations of demand in time and space.

In this talk, I will address the related challenges and present a two-layered hierarchical solution for the autonomous control of elastic DSP applications deployed in geo-distributed Cloud and Fog/edge environments. I will share some of our results, including the integration of application and infrastructure-level elasticity, and then try to identify and discuss a few open problems.

Call for paper

Nowadays, cloud systems and applications are moving from a Virtual Machine centric to a Container centric technology model. Containers are lightweight virtualization environment that exploit OS-virtualization features (also known as application virtualization) and is a technology that is changing the way cloud platforms and distributed applications are architected and managed. Containers enable micro-service software architectures, and are used to deploy and run enterprise, scientific and big data applications, to architect IoT and edge/fog computing systems, and by Cloud providers to internally manage their infrastructure and services.

The Container technology landscape is developing and expanding at the speed of light, however, we are far away from the maturity stage and there are still many research challenges to be solved. One of them is container orchestration, that demand for new solutions to select, deploy, monitor, and dynamically control the configuration of multi-container packaged applications in the cloud or in high performance computer systems. Another issue to be solved is security, for example the secure resource sharing and isolation to enable multi-tenancy without the need to leverage VMs boundaries. Other challenges are: the selection and configuration of host file system to improve performance; image layer caching to reduce deployment and scaling latency; container migration.

On the other side, as mentioned before, the use of containers span over many different application domains, and that raises new challenges to properly manage different type of workloads. The scientific objective of the workshop is to collect high quality contributions that: advance the state of the art of container technologies; propose new solutions for architecting high performance distributed and/or parallel container based systems; put forward resource management techniques like scaling and migration; enhance the state of the art in container security; show successful use of container technologies in fields like Big Data processing, Cloud Computing, Parallel computing, Distributed Computing and Internet of Things.

More precisely, the goal of the workshop is to share new findings, exchange ideas, discuss research challenges and report latest research efforts on the following subjects (not an exhaustive list):

• Performance modeling, Performance evaluation, Performance monitoring
• Workload characterization
• Orchestration models, mechanisms and policies for large-scale deployments
• Resource management at run time, autonomic computing
• Energy efficiency
• Security in container technologies and container-based systems
• Container images verification, software IPR etc. via Blockchain
• Containers to support smart IoT-based applications running in real-time
• Containers to support Edge and Fog computing
• Containers to support Big data systems
• Containers to support High Performance Computing
• Use cases and challenges in the areas of Cloud Computing/Applications, High Performance Computing, Parallel Computing, Distributed Computing, Big Data applications, IoT applications and Internet / Network Services
• Scheduling on heterogeneous multi-processing systems
• High-availability, geographic availability, and other non-functional requirements
• Repositories of container images and their non-functional requirements
• Containers and DevOps
• Component-based software engineering based on containers
• Container-based cloud networking (e.g. NFV)

Important dates

• Papers submission deadline: 27 May, 2018 (EXTENDED)
• Author notification: 25 June, 2018
• Informal camera-ready from the authors due: July 6, 2018
• Workshop formal camera-ready papers due: 02 October, 2018

Endorsements

• Technically co-sponsored by the Bonseyes project (www.bonseyes.com; a research and innovation action funded under EU's Horizon 2020 framework, grant agreement No 732204, and by SERI (Swiss State Secretariat for Education‚ Research and Innovation), under contract number 16.0159).
• Technically co-sponsored by the “Scalable resource-efficient systems for big data analytics” project (20140032), awarded by Knowledge Foundation, Sweden – namely the BigData@BTH project (http://www2.bth.se/bloggar/bigdata/)