The purpose of scenarios is to reduce the effort required to configure EdgeLab
The scenarios directory contains five custom scenarios. Within each of the scenarios directories
- 12e-t2-s0
- 12e-t2-s1
- 2e-t0-s1
- 6e-t1-s1
- 6e-t2-s1
There should be a config.aal and a run_scenario.sh. Note: in 6e-t2-s1 the run_scenario.sh script is in 6e-t1-s1.
The config.aal (agent animation list) contains a set of configuration settings used by magi.
The run_scenario.sh script will then call magi to start and stop certain modules:
- click
- general_viz
- http
- iperf
- mgen
- rtp
- simple_reorder
- targeted_loss
- tcpdump1
on the nodes specified in the config.aal file. The actions given in run_scenario.sh will use the action by the same name in the AALS/ directory (these are generally named start.aal and stop.aal for starting and stopping services. The agents_and_groups.aal and scenario_functions.sh are middleware that assists in allowing for easy to create scenarios.
There is also the modify_click/ directory which contains a framework for modifying click in an easy manner. The README gives additional details on using modify_click.
- Verify that MAGI is running on all nodes:
running:magi_status.py -e experiment_id -p project_id
should return:01-22 10:07:20.908 root INFO Received reply back from all the required nodes - ssh to your control node: (
control.experiment.project.isi.deterlab.net) cd /proj/edgect/scenarios/{SCENARIO}sudo ./run_scenario.sh
Additional Notes are given on the wiki