clientmonitor

Multi process monitor

This is a small helper library intended for monitoring multiple processes outputting the same data set, visualizing it in a ncurses table.

Example: Multiple scrapers processing different URLs logging their progress to one central monitoring entity.

Installation

pip install clientmonitor

Usage

This library consists of a server and a client. The Server accepts connections from clients and prints data received from the clients.

The clients expect the server to be running before connecting.

Running the server

Port and secret are mandatory, host is optional and only relevant if server and client are not being run on the same machine.

from clientmonitor.Server import Server

server = Server(6666, "secret", host="localhost")
server.set_columns([
  {
    "key": "key1",
    "label": "Label for Key 1"
  },
  {
    "key": "key2",
    "label": "Label for Key 2"
  }
])
server.start()

# CTRL-C to stop

The columns being set are the columns that are expected to be received as data from the clients and are what the server will print.

Running the client

import time
from clientmonitor.Client import Client

client = Client("client-1", 6666, "secret")
client.connect()

for i in range(0, 5):
  client.send({
    "key1": 100 + i,
    "key2": 200 + i,
  })

  time.sleep(5)

client.close()

Development

Pull request are welcome, pleas address them against the develop branch.

Publish

python -m build
python3 -m twine upload --repository testpypi dist/*