Edge information computing platform
The objective of the CLOUDLESS project is to develop an edge computing platform for the creation and deployment of open Data Spaces (International Data Spaces). The platform will follow a decentralized distributed architecture that offers event-based open discovery and interconnection services (data brokers) to different data consumers (data consumers) and data providers (data providers). The architecture will also offer secure Cloud/Edge infrastructures that allow efficient data analysis (data connectors) based on variables such as locality, economic cost, privacy or latency. The platform will be validated in different open data spaces such as citizen science, tourism or omics data (genomics and metabolomics).
Development of edge computing middleware that transparently optimizes across the continuum (Cloud/Edge)
Development of an edge computing platform for the management of open Data Spaces
Validation and dissemination of results on heterogeneous Cloud/Edge platforms including public Clouds, private clusters, IoT devices, and end-user devices (browsers, mobiles)
| Name | Description | Link |
|---|---|---|
| Lithops | The main serverless computing framework developed and extended during the project, featuring the 'Flexecutor' scheduler and enhanced WebAssembly support. | Repository |
| Lithops Applications (Benchmarks) | A suite of reference applications and benchmarks (Bioinformatics, Geospatial, Metabolomics) used for KPI and performance validation. | Repository |
| SpotWhisk | SpotWhisk is a FaaS platform that replaces standard computational units with spot instances to achieve significant cost savings. | Repository |
| Water Consumption Pipeline | A serverless pipeline designed to measure and analyze water consumption across different geographic regions. | Repository |
| DataPlug | An open-source library for cloud data management that enables on-the-fly dynamic partitioning of unstructured and semi-structured data. | Repository |
| DataCockpit | A visual tool (Python widget) that enables simple and intuitive dataset partitioning through a graphical interface without writing code. | Repository |
| ML-Pipeline-Optimizer | Repository including the platform and results of the publication 'Intelligent Optimization of Distributed Pipeline Execution in Serverless Platforms: A Predictive Model Approach'. | Repository |
| Dynamic Frequency Based Fingerprinting Attacks against Modern Sandbox Environments | An attack framework that identifies cloud containers and sandboxes via CPU frequency fingerprints and proposes mitigation through noise injection. | Repository |
| OpenNebula Appliances | Official repository containing the appliances developed to deploy Lithops and MinIO on clusters managed by OpenNebula (K8s/LXC). | Repository |
| Title | Link |
|---|---|
| Dataplug: Unlocking extreme data analytics with on-the-fly dynamic partitioning of unstructured data | Open |
| Exploiting Inherent Elasticity of Serverless in Algorithms with Unbalanced and Irregular Workloads | Open |
| Intelligent Optimization of Distributed Pipeline Execution in Serverless Platforms: A Predictive Model Approach | Open |
| Exploring Secure and Efficient Temporary Data Sharing between co-located Kubernetes Containers | Open |
| Burst Computing: Quick, Sudden, Massively Parallel Processing on Serverless Resources | Open |
| Optimizing WebAssembly Garbage Collection in Go: Performance Insights, Tuning Tips, and Batch Execution Strategies | Open |
| Dynamic Selection and Detection of Spreading Factors and Channels for End-Node Devices of LoRa Networks | Open |
| Let It Unthread: The Good, The Bad and The Ugly within WebAssembly Portable Multithreading | Open |
| The Hidden Dangers of Public Serverless Repositories: An Empirical Security Assessment | Open |
| Dynamic Frequency-Based Fingerprinting Attacks against Modern Sandbox Environments | Open |
| Rethinking the mobile edge for vehicular services | Open |
| Energy-Efficient Task Computation at the Edge for Vehicular Services | Open |
PyRun is defined as the world's first "Serverless Python Studio", a deep-tech platform that acts as an intelligent bridge between local code and the immense power of the cloud. Its critical importance lies in eliminating the "complexity tax": currently, data scientists and AI engineers lose up to 60% of their time configuring complex infrastructures instead of innovating. PyRun solves this by democratizing supercomputing; it allows any Python user, without DevOps knowledge, to run their scripts on clouds like AWS or IBM with a single click, while the platform automatically manages the entire server lifecycle, ensuring that technical talent focuses purely on creating value rather than "connecting cables".
| Project title | CLOUDLESS: Edge information computing platform |
| Coordinator | Dr. Pedro García López (URV) |
| Partners |
Universitat Rovira i Virgili
OpenNebula Telefónica Research Alterna Tecnologías Universidad de Zaragoza |
| Duration | 01/01/2023 - 30/06/2025 |
| Overall budget | 2.351.607,14 € |
| Funding | Programa UNICO I+D Cloud, en el marco del Plan de Recuperación, Transformación y Resiliencia -Financiado por la Unión Europea- Next Generation EU |
| Dissemination materials | Brochure - Poster |