NSF-CICI

Useful References:

ESNET DTNs

ESNET Switch Monitoring

Experiment Setup Commands:

Setting up a Systemd Service:

# Create a systemd unit file
sudo cp <unit file> /lib/systemd/system/
sudo chmod 644 <unit file> 
sudo systemctl daemon-reload #Reread the modified unit file 
sudo systemctl enable <unit file> #Starts service at boot
(reboot) 
sudo systemctl start <unit file> 
sudo systemctl restart <unit file> #Restart service

To Check if background logging and Globus transfers are running:

sudo systemctl status  collectstat.service  collectnet.service  collectinterrupt.service globus.service 

Setting up Journalctl logging:

Edit /etc/systemd/journal.conf:

storage = persistent

Make a new directory to persistently store Journald logs, otherwise logs are stored in /run/log/journal which volatile on reboots.

mkdir /var/log/journal

To check Journald logs:

sudo journalctl --list-boots #shows previous boots
sudo journalctl -u globus.service -b <boot#>

Hping Attacks:

echo "<script location>" | sudo at <time> tomorrow

Data on Kelewan:

* Collectl: /var/log/collectl/
* Procfs: /home/cmao/Repos/nsf-cici/data/procfs/

Workflow:

1. Copy data files from Kelewan to local (/Repos/nsf-cici/data/<experiment_name>) * Collectl * ProcFS/raw
2. Clean ProcFS raw files with clean_logs.py
   • Set File Date
   • Set STATDIR
   • Set NETDIR
   • Set OUTDIR </procfs/clean>
3. Run htm_procfs.py on each x_clean_csv file
   • Set FILENAME
   • Set INPUT_DIR
   • Set OUT_DIR
   • Check Model Parameters
4. Run plot_htm_scores.py
   • Set DATE of the file