On 4th of April, 2022 the Data & Web Science Laboratory of Aristotle University of Thessaloniki (AUTH) with the support of RAINBOW successfully organized an online workshop on the topic of “Processing Data in the Fog. The example of the RAINBOW Fog Computing platform”. During this workshop various members of the research teams of AUTH, UBITECH and UCY presented a series of research challenges in Fog and Edge environments from the scope of data management and distributed data processing, as well as the innovative solutions developed by RAINBOW to address them.
Figure 1: Opening & Welcome / Speakers: Prof. Athena Vakali (AUTH) & Assoc. Prof. Anastasios Gounaris (AUTH)
In the first session Alexandros Vasileiou (UBITECH) presented an overview of the RAINBOW platform, its high-level architecture and how it aims to empower IoT services operators and cross cloud app developers along with cloud infrastructure providers to capitalize on the potential of Fog Computing.
Figure 2: Session 1 / Speaker: Alexandros Vasileiou (UBITECH)
In the following session Theodoros Toliopoulos (AUTH) described a distributed data storage architecture and its implementation in RAINBOW’s storage agent along with two novel data placement techniques developed to tackle the challenges of Fog environments, such as resource limitations and Fog-nodes instability.
Figure 3: Session 2 / Speaker: Theodoros Toliopoulos (AUTH)
George Vlahavas (AUTH) continued with the presentation of a series of benchmarking experiments aiming to analyse the performance and behaviour of popular distributed in-memory DBMS, such as Redis and Apache Ignite, in dynamic fog environments.
Figure 4: Session 3 / Speaker: George Vlahavas (AUTH)
Dr. Demetris Trihinas (UCY) presented StreamSight, a streaming analytics framework supporting multiple big data engines, such as Apache Spark and Apache Storm, and offering an SQL-like query model along other Fog optimization hints.
Figure 5: Session 4 / Speaker: Dr. Demetris Trihinas (UCY)
The fifth session given by Anna-Valentini Michailidou (AUTH) provided an overview of task and data and placement optimization algorithms able to respond to the requirements of a Fog Computing data management and analytics stack, such as resources and network heterogeneity, availability and privacy constraints, latency, etc.
Figure 6: Session 5 / Speaker: Anna-Valentini Michailidou (AUTH)
In the final session Moysis Symeonides (UCY) presented the Fogify Fog emulation framework, that aims to address many of the Fog design and deployment challenges, and how it was used in RAINBOW for performance evaluation and quantitative experiments in the project use cases.
Figure 7: Session 6 / Speaker: Moysis Symeonides (UCY)
The workshop attracted an audience of 48 participants with 2/3 being from academia (computer science) and ¼ coming from the industry. According to an anonymous feedback survey the majority of the attendees responded that were very much satisfied with both the organization and the quality of the presentations and content of the workshop. More over its especially interesting to note that the participants rated the potential overall impact of the work done within RAINBOW in Fog/Edge computing from ‘significant’ to ‘strong’.