Edge computing and the Internet of Things (IoT) expand the range of possibilities for the use of computing resources. Smart factories, smart cars, or worker accident detection are examples of IoT use cases that are enabled by the Edge computing paradigm. Many scenarios require combining Edge computing with Cloud computing to allow leveraging high-performance computing resources in the Cloud. This combination allows using any device, starting from the Edge all the way until (and including) the Cloud — this meld of Edge and Cloud computing is often referred to as the Edge-to-Cloud continuum. One of the key features of the Edge-to-Cloud continuum is elasticity, i.e., adapting resource, quality, and cost characteristics of a workload according to the current needs. In this hackathon we will focus on the resource elasticity aspect.
While there are many potential use cases for the Edge-to-Cloud continuum, their realization requires adequate testing tools to verify the solutions. Testing the elasticity of an application can be particularly challenging.
Elasticity is normally guided by Service Level Objectives (SLOs), which define the bounds within which a service should operate. A simple and common example is the following: an SLO may specify that a REST API service should operate at an average CPU usage between 50% – 70% – if it exceeds this target range, the service should scale out (i.e., more instances should be deployed) and if it goes below this range, the service should scale in (i.e., one or more instances should be stopped).
In this hackathon participants will create a load generator that can be used to push deployed services to their limits in order to test their SLOs and scaling. In addition, they will create a deployment and SLO(s) on the RAINBOW platform to demonstrate your work.
For detailed instructions on how to use the RAINBOW platform, please see its documentation: https://rainbow-h2020.readthedocs.io
All code created as part of this hackathon must be released as open source (e.g., on GitHub or GitLab) and must be licensed under the Apache-2.0 license.
The hackathon will be held online on December 1, 2022. The agenda is shown below:
| 10:00 – 10:15 | Welcome and formation of teams |
| 10:15 – 11:00 | Overview of Edge/Fog Computing |
| 11:00 – 11:15 | Presentation of the Hackathon challenge |
| 11:15 – 15:00 | Teams work on their solutions |
| 15:00 – 16:15 | Presentation of solutions (10 min presentation + 5 min Q&A per team) |
| 16:15 – 16:30 | Break |
| 16:30 – 16:45 | Award presentation |
In advance registration is required at:
Contact members:
Marco Rapelli (Politecnico di Torino) –> marco.rapelli@polito.it
Thomas Pusztai (TU Wien) –> t.pusztai@dsg.tuwien.ac.at
Heini Bergsson Debes (DTU) –> heib@dtu.dk
