- Run "python generate.py -h" to see arguments. Run generate.py to generate traffic data first. Defaults works for simple topology with 2 servers. Output will be a file, use this file for the next step.
- Run "python prepare.py <input_file>" to divide the data for each individual host.
- Update host IPs in sendFiles.sh, backgroundTraffic.sh, servers.json, kill_listeners.sh
- Send the output file to hosts. Run the script
./sendFile.sh
to send it. - Execute
./backgroundTraffic.sh
to start background traffic on all servers - Execute
./kill_listeners.sh
to kill the listening servers.
- generate.py: seed, load, num_spines, num_leaves, num_servers_per_rack, end_time, outfile
- time_walking.py dst, num_servers
- max_time in time_walking.py