About

graph LR
Webpage[Web page] <--Flask--> Server[Server]
Server <-- Push agent and execute --> Client((Client))
Server[Server] <----> Client2((Client2))
Server[Server] <----> Client3((Client3))

Simple monitoring script. It will push an agent (agent.py) on remote servers to get info about ressources (CPU, Disk, Memories..) And display it on a webpage.

Screenshot

Install

For Debian 12 (and more...)

On server :

apt install python3-flask python3-paramiko python3-psutil

On nodes (if needed):

apt install python3-psutil

For Debian < 12

pip3 install psutil flask paramiko

On nodes :

pip3 install psutil

If you prefer (recommended), you can run in an python env

apt install python3-venv

python3 -m venv Pymonit --system-site-packages
. /Pymonit/bin/activate
pip3 install psutil flask paramiko

Configuration

Clients

Add your clients in the python list in config.py:

client = ['192.168.0.1','server.local','server.mydns']

Generate keys for connection

Becarrefull, avoid using a root user to start application and connect to client. Use a dedicated user.

On server, generate a pair of key :

ssh-keygen -b 2048 -t rsa

And push the public key in your clients :

ssh-copy-id -i <home_user>/.ssh/id_rsa.pub <user>@<client>

Configuration on application

You have to report these info on config.py :

class ssh_conf:
       port = 22
       username = '<user>'
       password ='<home_user>/.ssh/id_rsa.pub'

For testing purpose, but still not recommended, you can use a password in configuration file.

class ssh_conf:
       port = 22
       username = '<user>'
       password ='<password>'

Flask configuration

Configure the flask configuration according to your needs (port,host,...) in config.py

class flask_conf:
       port = 8090
       host = "0.0.0.0"
       thread = True
       debug = False
       reloader = False

NB: Let thread = True, else this will take a time to switch from node to another node

Path configuration

class paths :
    basedir = os.path.abspath(os.path.dirname(__file__))
    agent_path = basedir+'/agent.py'
    remote_path = '/var/tmp/script.py'

remote_path : the location where the agent will be pushed and execute. Take care that your used is able to write and execute in this location.

Systemctl (optional)

To easy start, stop and restart application, you can use the monit.service template.

Replace the path according to your system in :

[Unit]
Description= Monit service
After=multi-user.target
[Service]
Type=simple
Restart=always
ExecStart=/usr/bin/python3 /application/path/app.py
[Install]
WantedBy=multi-user.target

Copy the file in your systemd folder :

cp ./monit.service /etc/systemd/system/

Then you can start it :

systemctl start monit.service

To get it automaticaly start with your system:

systemctl enable monit.service

Change log

v1 :

• Create Agent:

  • CPU detail
  • Memory detail
  • Swap detail

• Agent communication :

  • SSH using Python
  • Push and Run agents
  • Retrieve info

• Interface:

  • Creation
  • Bootstap
  • Delete

v1.1

• Adding Disk usage info

  • Get partitions
  • Get usage

• Adding Network info:

  • Get interfaces
  • Get info (IP,netmask)

v2:

• Create Live view of SWAP,CPU,Memory usage
• Adding Chart to interface

v2.1 On going :

• Add Network usage
• Add boot time
• Update dynamically Progress Bar

Next:

• Retrieve Process runnning (top ?)
• Add alerting (?)
  • Create admin page for threshold and alert
  • Manage alert communication (Mail, SMS...)