TTFB Measurements in guifi.net
$pip install virtualenv
$git clone https://github.com/emmdim/TTFB_ML.git && cd TTFB_ML
$virtualenv venv
$source venv/bin/activate
$pip install --upgrade pip
$pip install -r requirements.txt
DURING THE FOLLOWING STEP SUDO ACTIONS WILL PROMPT FOR PASSWORDS
Copy your private key to the TTFB_ML folder name as key and your public key in the same place named as key.pub
Configure the clients addresses in the clients variable inside the fabfile.py. The addresses need to be in the format user@ip
The same needs to be done for the servers
Add the necessary entries to ssh-config for each used client and server.
Install your public key to the clients executing:
fab clients_upload_pubkey
Since the servers are more to be used with more care you have to install by yourself your public key to them.
The ssh_config file is an ssh configuration file that contains information for that can be leveraged when connecting with ssh key. Please modify in this step according to your needs.
Then, prepare the environment for the servers and the clients:
fab servers_deploy
fab clients_deploy
All the functions from this point on can be modified to fill the needs. The existing version creates necessary directories, downloads necessary packages, downloads the project code, and installs a virtualenv where necessary.
The part of the modules of the code concerning the servers are placed in the servers folder. Similarly for the clients.
The clients_pull and the servers_pull update the code on the clients and the servers respectively.
The rest of the functions in the fabfile can be customized as needed.
The getEET.py module provides the getEET function that returns a UNIX timestamp and can be used to synchronize servers and clients.
The requests.py module contains the logic for the periodical requests from the clients to the servers.
The check-running.sh script can be installed as a cronjob to monitor if the requests.py module is alive and take action otherwise.
The monitor.py module contains the logic for the monitoring of the squid proxy and other metrics, like RX,TX etc.
The check-running.sh script can be installed as a cronjob to monitor if the requests.py module is alive and take action otherwise.