get process of a running browser

[shobu13]

I need to monitor energy consumption of a browser (for exemple chromium) testing multiples pages of the database. I want to use django to runnings my tests concurently using playwright to automate the browser testing (i've already use it a little using the JS api), but the tool i use to monitor energy consumption watch a cgroup where the runing process of the browser need to be added. For this purpose, i've checked on the internet if it's possible to get the PID of a browser using PW and it's seem possible using

  const browserServer = await playwright.chromium.launchServer()
  const pid = await browserServer.process().pid
  console.log(pid)

https://stackoverflow.com/questions/62376437/get-pid-of-a-browser-launched-with-playwright

But i don't found a way to retrieve this information in playwright-go due to the fact that pw.Chromium don't have a function LauchServer and i've don't found an equivalent.

[Zunaira518]

@mxschmitt Is there any resolution for this?
I'm encountering the same problem where Chrome launches for every API call. Despite using a defer() function on closure for page, context, and browser, it still results in zombie processes that consume all the memory over time. It would be helpful to have a function that returns the process IDs of the browsers started by Playwright, or an alternative solution to manage this issue.

> root@saviynt-inspector-main-43-test-c4b8b5d66-lfzp2:/app# ps -eaf
> UID          PID    PPID  C STIME TTY          TIME CMD
> root           1       0  0 14:09 ?        00:00:05 /app/saviynt-inspector
> root          22       1  0 14:09 ?        00:00:00 /bin/sh /root/.cache/ms-playwright-go/1.40.1/playwright.sh run-driver
> root          25      22  0 14:09 ?        00:01:07 /root/.cache/ms-playwright-go/1.40.1/node /root/.cache/ms-playwright-go/1.40.1/package/lib/cli/cli.js run-driver
> root         105       1  0 14:09 ?        00:00:00 [chrome] <defunct>
> root         106       1  0 14:09 ?        00:00:00 [chrome] <defunct>
> root         239       1  0 14:09 ?        00:00:00 [chrome] <defunct>
> root         240       1  0 14:09 ?        00:00:00 [chrome] <defunct>
> root         377       1  0 14:10 ?        00:00:00 [chrome] <defunct>
> root         378       1  0 14:10 ?        00:00:00 [chrome] <defunct>
> root         509       1  0 14:10 ?        00:00:00 [chrome] <defunct>
> root         510       1  0 14:10 ?        00:00:00 [chrome] <defunct>
> root         642       1  0 14:11 ?        00:00:00 [chrome] <defunct>
> root         643       1  0 14:11 ?        00:00:00 [chrome] <defunct>
> root         772       1  0 14:11 ?        00:00:00 [chrome] <defunct>
> root         773       1  0 14:11 ?        00:00:00 [chrome] <defunct>

[mxschmitt]

@Zunaira518 are you running inside Docker? Make sure to use --init as per the Docker best practises. This is usually the reason why defunct processes are staying there.

[Zunaira518]

@mxschmitt Adding --init fixed the issue. Thankyou for quick reply :)