Pyppeteer: Browser closed unexpectedly in AWS Lambda - aws-lambda

Question

I'm running into this error in AWS Lambda. It appears that the devtools websocket is not up. Not sure how to fix it. Any ideas? Thanks for your time.
Exception originated from get_ws_endpoint() due to websocket response timeout https://github.com/pyppeteer/pyppeteer/blob/ad3a0a7da221a04425cbf0cc92e50e93883b077b/pyppeteer/launcher.py#L225
Lambda code:
import os
import json
import asyncio
import logging
import boto3
import pyppeteer
from pyppeteer import launch
logger = logging.getLogger()
logger.setLevel(logging.INFO)
pyppeteer.DEBUG = True # print suppressed errors as error log
def lambda_handler(event, context):
asyncio.get_event_loop().run_until_complete(main())
async def main():
browser = await launch({
'headless': True,
'args': [
'--no-sandbox'
]
})
page = await browser.newPage()
await page.goto('http://example.com')
await page.screenshot({'path': '/tmp/example.png'})
await browser.close()
return {
'statusCode': 200,
'body': json.dumps('Hello from Lambda!')
}
Exception:
Response:
{
"errorMessage": "Browser closed unexpectedly:\n",
"errorType": "BrowserError",
"stackTrace": [
" File \"/var/task/lambda_handler.py\", line 23, in lambda_handler\n asyncio.get_event_loop().run_until_complete(main())\n",
" File \"/var/lang/lib/python3.8/asyncio/base_events.py\", line 616, in run_until_complete\n return future.result()\n",
" File \"/var/task/lambda_handler.py\", line 72, in main\n browser = await launch({\n",
" File \"/opt/python/pyppeteer/launcher.py\", line 307, in launch\n return await Launcher(options, **kwargs).launch()\n",
" File \"/opt/python/pyppeteer/launcher.py\", line 168, in launch\n self.browserWSEndpoint = get_ws_endpoint(self.url)\n",
" File \"/opt/python/pyppeteer/launcher.py\", line 227, in get_ws_endpoint\n raise BrowserError('Browser closed unexpectedly:\\n')\n"
]
}
Request ID:
"06be0620-8b5c-4600-a76e-bc785210244e"
Function Logs:
START RequestId: 06be0620-8b5c-4600-a76e-bc785210244e Version: $LATEST
---- files in /tmp ----
[W:pyppeteer.chromium_downloader] start chromium download.
Download may take a few minutes.
0%| | 0/108773488 [00:00<?, ?it/s]
11%|█▏ | 12267520/108773488 [00:00<00:00, 122665958.31it/s]
27%|██▋ | 29470720/108773488 [00:00<00:00, 134220418.14it/s]
42%|████▏ | 46172160/108773488 [00:00<00:00, 142570388.86it/s]
58%|█████▊ | 62607360/108773488 [00:00<00:00, 148471487.93it/s]
73%|███████▎ | 79626240/108773488 [00:00<00:00, 154371569.93it/s]
88%|████████▊ | 95754240/108773488 [00:00<00:00, 156353972.12it/s]
100%|██████████| 108773488/108773488 [00:00<00:00, 161750092.47it/s]
[W:pyppeteer.chromium_downloader]
chromium download done.
[W:pyppeteer.chromium_downloader] chromium extracted to: /tmp/local-chromium/588429
-----
/tmp/local-chromium/588429/chrome-linux/chrome
[ERROR] BrowserError: Browser closed unexpectedly:
Traceback (most recent call last):
  File "/var/task/lambda_handler.py", line 23, in lambda_handler
    asyncio.get_event_loop().run_until_complete(main())
  File "/var/lang/lib/python3.8/asyncio/base_events.py", line 616, in run_until_complete
    return future.result()
  File "/var/task/lambda_handler.py", line 72, in main
    browser = await launch({
  File "/opt/python/pyppeteer/launcher.py", line 307, in launch
    return await Launcher(options, **kwargs).launch()
  File "/opt/python/pyppeteer/launcher.py", line 168, in launch
    self.browserWSEndpoint = get_ws_endpoint(self.url)
  File "/opt/python/pyppeteer/launcher.py", line 227, in get_ws_endpoint
    raise BrowserError('Browser closed unexpectedly:\n')END RequestId: 06be0620-8b5c-4600-a76e-bc785210244e
REPORT RequestId: 06be0620-8b5c-4600-a76e-bc785210244e Duration: 33370.61 ms Billed Duration: 33400 ms Memory Size: 3008 MB Max Memory Used: 481 MB Init Duration: 445.58 ms

Answer 1

I think BrowserError: Browser closed unexpectedly is just the error you get when Chrome crashes for whatever reason. It would be nice if pyppeteer printed out the error, but it doesn't.
To track things down, it's helpful to pull up the exact command that pyppeteer runs. You can do that this way:
>>> from pyppeteer.launcher import Launcher
>>> ' '.join(Launcher().cmd)
/root/.local/share/pyppeteer/local-chromium/588429/chrome-linux/chrome --disable-background-networking --disable-background-timer-throttling --disable-breakpad --disable-browser-side-navigation --disable-client-side-phishing-detection --disable-default-apps --disable-dev-shm-usage --disable-extensions --disable-features=site-per-process --disable-hang-monitor --disable-popup-blocking --disable-prompt-on-repost --disable-sync --disable-translate --metrics-recording-only --no-first-run --safebrowsing-disable-auto-update --enable-automation --password-store=basic --use-mock-keychain --headless --hide-scrollbars --mute-audio about:blank --no-sandbox --remote-debugging-port=33423 --user-data-dir=/root/.local/share/pyppeteer/.dev_profile/tmp5cj60q6q
When I ran that command in my Docker image, I got the following error:
$ /root/.local/share/pyppeteer/local-chromium/588429/chrome-linux/chrome # ...
/root/.local/share/pyppeteer/local-chromium/588429/chrome-linux/chrome:
error while loading shared libraries:
libnss3.so: cannot open shared object file: No such file or directory
So I installed libnss3:
apt-get install -y libnss3
Then I ran the command again and got a different error:
$ /root/.local/share/pyppeteer/local-chromium/588429/chrome-linux/chrome # ...
[0609/190651.188666:ERROR:zygote_host_impl_linux.cc(89)] Running as root without --no-sandbox is not supported. See https://crbug.com/638180.
So I needed to change my launch command to something like:
browser = await launch(headless=True, args=['--no-sandbox'])
and now it works!

Answer 2

Answering my own question.
Finally I was able to run Pyppeteer(v0.2.2) with Python 3.6 and 3.7 (not 3.8) after I bundled chromium binary in a lambda layer.
So in summary, it appears to work only when its configured to run with user provided chromium executable path and not with automatically downloaded chrome. Probably some race condition or something.
Got Chromium from https://github.com/adieuadieu/serverless-chrome/releases/download/v1.0.0-41/stable-headless-chromium-amazonlinux-2017-03.zip
browser = await launch(
headless=True,
executablePath='/opt/python/headless-chromium',
args=[
'--no-sandbox',
'--single-process',
'--disable-dev-shm-usage',
'--disable-gpu',
'--no-zygote'
])
Issue posted on repo https://github.com/pyppeteer/pyppeteer/issues/108

Answer 3

I have been trying to run pyppeteer in a Docker container and ran into the same issue.
Finally managed to fix it thanks to this comment: https://github.com/miyakogi/pyppeteer/issues/14#issuecomment-348825238
I installed Chrome manually through apt
curl -sSL https://dl.google.com/linux/linux_signing_key.pub | apt-key add -
echo "deb [arch=amd64] https://dl.google.com/linux/chrome/deb/ stable main" > /etc/apt/sources.list.d/google-chrome.list
apt update -y && apt install -y google-chrome-stable
and then specified the path when launching the browser.
You also have to run it in headless and with args "--no-sandbox"
browser = await launch(executablePath='/usr/bin/google-chrome-stable', headless=True, args=['--no-sandbox'])
Hope this will help!

Answer 4

If anybody is running on Heroku and facing the same error:
Add the buildpack : The url for the buildpack is below :
https://github.com/jontewks/puppeteer-heroku-buildpack
Ensure that you're using --no-sandbox mode
launch({ args: ['--no-sandbox'] })

Answer 5

Make sure all the necessary dependencies are installed. You can run ldd /path/to/your/chrome | grep not on a Linux machine to check which dependencies are missing.
In my case, i get this:
libatk-bridge-2.0.so.0 => not found
libgtk-3.so.0 => not found
and then install dependencies:
sudo apt-get install at-spi2-atk gtk3
and now it works!