Add ability to monitor processes not started by cardamon

[seal]

Example used:

• Next.js application that calculates Prime numbers ( client side, CPU heavy )
  
• Node.js "puppeteer" application that navigates to the UI, and generates 10000 prime numbers

const puppeteer = require('puppeteer');
(async () => {
    try {
        const browser = await puppeteer.launch({
            headless: true, // Keep as true
        });
        const page = await browser.newPage();
        await page.goto('http://127.0.0.1:3001');
        await page.type('#limit', '1000000');
        // generate prime numbers button
        await page.click('.bg-blue-500.text-white.p-2.rounded');
// etc
    } catch (error) {
        console.error('An error occurred:', error);
    }
})();

The idea of this example is to generate a "CPU Leak" / memory leak on the client side

The problem with this is:

• cardamon can only record PID's of processes it spawns
• Puppeteer spawns chromium ( browser) not as a child process but as a detached process

This means after recording the CPU usage of our node.js program, we get a CPU usage of 0

Proposed changed:

• Add the ability to send Process ID's to cardamon after it's started

One implementation of doing this would be as follows:

• Create a http-server via cardamon
• Await a post request with process ID's to start the scenario

Example node.js code:

const puppeteer = require('puppeteer');
const request = require('request');
(async () => {
    try {
        console.log('Starting puppeteer');
        const browser = await puppeteer.launch({
            headless: true, // Keep as true
        });
        request.post(
            'http://127.0.0.1:XX/api/pid',
            { json: { name: 'browser_click', pid: browser.process().pid } },
            function(error, response, body) {
                if (!error && response.statusCode == 200) {
                    console.log(body);
                }
            }
        );

        const page = await browser.newPage();

        await page.goto('http://127.0.0.1:3001');

        await page.type('#limit', '1000000');

        // generate prime numbers button
        await page.click('.bg-blue-500.text-white.p-2.rounded');


        // This will timeout at 30s 
        await page.waitForNavigation({ waitUntil: 'networkidle0' });

        console.log('Actions completed successfully');

        await browser.close();
    } catch (error) {
        console.error('An error occurred:', error);
    }
})();

Example cardamon.toml config:

[[processes]]
name = "pupp"
up = "node puppeteer.js"
[processes.redirect]
to = "parent"
[processes.process]
type = "baremetal"
[process.pid]
hang = true
name = "browser_click"

The idea here is to hang the starting of logging / metrics until the "pupp" process has sent a PID