Queueing simulator
This repository contains a simulator for queueing network designed for quick experimentations. It supports both FCFS and PS strategies, stateful transitions between queues, etc.
Current experiments
Fog admission control
A queueing network to simulate a Fog-acceptance module to control application latency. See [1] for further description of the scenario.
To run it, set the parameters in src/fog_cloud_sim
, then run cargo run fog [mode]
where [mode]
is one of:
blind
for a blind admission-control modulelfu
for the oracle-based admission-controllru
for the LRU-AC implemented with an actual LRUabf
for the LRU-AC implemented with an ABF cache
To derive the parameters of the Admission-Control module, a Jupyter notebook is available in helpers/Fog admission control optimization
. You can use it by installing Jupyter.
Note that this notebook must be ran with a Python 2 kernel
[1] : Enguehard, Marcel, Giovanna Carofiglio and Dario Rossi. “A Popularity-Based Approach for Effective Cloud Offload in Fog Deployments.” 2018 30th International Teletraffic Congress (ITC 30)
Autoscaling
To be completed soon...