I am working on an Edge Framework which aims to execute tasks/processes in real-time and to offload the tasks to other neighboring Edge Servers in case the current server does not have enough resources, or likely to miss the deadline.
There are several architecture available from e.g. OpenFog Consortium or EdgeX Foundry. Additionally, simulators such as iFogSim, EdgeCloudSim (which are based on CloudSim), EmuFog, or YAFS exist. However, as far as I understood, they rely on best effort rather than real-time reaction.
Are there any existing solutions in the same direction, i.e. real-time execution in a distributed network?
Thank you,
Hello! You can check my paper which proposes serverless architecture for workflow/task execution. It doesn't use Edge Servers as a workers, but the I think the other parts of the architecture could be useful to you. Let me know what you think.
Hi Volkan Gezer ! I have been working on (re)configuring latency-sensitive applications across cloud and edge resources. In my case, I developed solutions atop Omnet++. I strongly believe that my research can assist you with your problem. In my proposed mechanisms, I applied methods from queueing theory, reinforcement learning, graph theory, and so on.
Regarding infrastructures, I would like to mention some: FIT/IoT-LAB (shared edge infrastructure) and R-Pulsar (http://rpulsar.rdi2.rutgers.edu/). I hope to provide you with helpful information. If you want to talk about your problem, do not hesitate to contact me.
looks similar to my previous work, you can find my paper here on researchgate, I used the following simulator:
https://github.com/CharafeddineMechalikh/PureEdgeSim
Thank you for the answers. I am continuously checking the tools out you mentioned.
- Badges
- Science topic
- Similar topics
- Design
- Architecture