How to get all processes in a display from Xvfb? - google-chrome

I have a program that is launching ChromeDrivers, which launches Chrome browsers + later attempts to close both after doing some task (using Selenium). But often times my program can't kill off the ChromeDriver/Chrome browser. When I try to kill the ChromeDriver the Chrome browser + all it's other child processes aren't killed off.
I have tried to look at /proc/x/environ to determine if I can extract the DISPLAY of the process, but found that no such environment variable was set for the browser + child processes.
Is there any other way to detect all processes in a specific Xvfb display and kill them all?

This looks quite promising if you only have one child process:
xvfb-run sleep 60 &
pid_xvfb=$!
kill $(ps -o pid= --ppid $pid_xvfb)

Related

Prompted by "Google Chrome didn't shut down correctly.... "

Using Google Chrome 38.0.2125.111 m in a Windows 7 environment. It is launched by a batch file & set to go to a specific (internal) URL. The Windows environment is going to be locked down, permitting only a handful of options available to the user.
Chrome is being used because IE will be accessing a different site. Without going into boring detail, using both sites on a single browser allowed a user to access an account that wasn't theirs. Both browsers will be launched using one of two batch files. The Chrome batch file supplies the username & password to bypass a login screen on that application.
To keep from opening multiple windows in Chrome, the following batch file is used:
taskkill /f /t /im chrome.exe > NUL
ping -n 3 127.0.0.1 > NUL
START /max chrome "http://myinternalwebsite?uid=x&pwd=y"
Taskkill is used to close chrome (/f = force /t = terminates process & any child ones /im = image name).
If the Chrome session is closed (using the X at the upper right corner of the screen), the batch file works fine. However, if there is already a chrome instance running, we get the yellow bar with the frowning folder & a message saying:
Google Chrome didn't shut down correctly. To reopen the pages you had
open, click Restore.
I have tried several command line options when launching Google Chrome, but none will keep the message from displaying. Ones I have tried are:
--disable-restore-session-state
--disable-session-crashed-bubble
--incognito
among others. I've experimented with some of the settings in Chrome, but none have (yet) worked. The --incognito option works - sort of. It keeps blocking plugins, one of which is Silverlight - one of the chief components for the website to function. Even though it does get rid of the message, it leaves the site not functioning properly.
The intent on using Taskkill to flush any existing Chrome sessions was to prevent the user opening up a dozen different Chrome instances of the same thing.
Is there any way to keep a user from opening multiple instances of Chrome that won't leave Chrome thinking it didn't close properly? Given the user restrictions, I'm not sure that a script/batch file to edit the Preferences file in the Chrome folder will work. An extension may be an option, but a third party application will not be.
Would appreciate any thoughts or assistance.
Thanks!
Chromium writes exit status into Preferences file. When it starts it writes Preferences file with:
"exited_cleanly": false
and upon clean exit writes Preferences file with:
"exited_cleanly": true
To prevent message about unclean shut down make sure that you have "exited_cleanly": true written in your Preferences file
On my Linux kiosk running on Raspberry Pi this sed command line added to /etc/rc.local fixes issue permanently:
sed -i 's/exited_cleanly\":\ false/exited_cleanly\":\ true/g' /home/pi/.config/chromium/Default/Preferences

How to kill a zombie process which always initiated whenever geany does

I am using Geany for editing a large text data in Ubuntu (600MB or so). But after a while, a zombie process starts whenever I start Geany and it couldn't load the file so that I edit the content. It took 100% of my CPU while Geany runs. I try to kill the zombie process with the following:
kill -HUP `ps -A -ostat,ppid,pid,cmd | grep -e '^[Zz]' | awk '{print $2}'`
But once I start the application again, the zombie process starts automatically. Also tried logout.
What can I do to kill the zombie process once and for all? Thanks!
You can't kill a zombie process since it's already dead.
On Unix and Unix-like computer operating systems, a zombie process or
defunct process is a process that has completed execution (via the
exit system call) but still has an entry in the process table: it is a
process in the "Terminated state".
(from Wikipedia)
It's simply an entry in the process table with no associated process. It exists because the spawning (parent) process has yet to collect the return status (via wait()). Other than that it will consume no resources.
So I suspect the parent process is either busy or not working properly. I would first of all try to identify that process (via the PPID column in ps, for example)
EDIT: I note there's a geany issue raised/resolved around this

CGI Bash script to spawn daemon process

I am working on a project to stream HDTV from a personal computer to devices supporting HTTP Live Streaming (think iOS devices and some android). I have the formatting of the video and the streaming aspects down. What I am now trying to implement is an easy way to change the channel remotely.
My current method involves connecting via SSH to kill the old stream and begin a new stream. This works, but isn't pretty. I want something my Mom or girlfriend could use. I decided I would build an HTML5 app that would issue the channel switching over CGI scripts. I currently have a parent process with a form that calls a child process to decide if the stream is running and then a subchild process to actually tune the stream.
As I am streaming live video from my computer I need the subchild process to run indefinitely. Unfortunately it seems that when my parent process is finished the background process started in the subchild process terminates.
I have tried a simple &, using nohup, setsid, and daemon. daemon runs cleanest but still terminates when the parent finishes. even with a -r flag. I'll place my code below and maybe someone will have an idea on how I could implement this or a better way to achieve the same thing? Thanks! (oh and i know killing vlc is not a pretty way to kill the stream, if you have a better way i'm all ears)
parent invoking child:
----------------------
./ChangeChannel.sh $channel #passed from form submission
child (ChangeChannel.sh):
-------------------------
#!/bin/bash
directory=./Channels/
newchannel=$1
if [ $(pidof vlc) ]
then
sudo kill $(pidof vlc)
fi
daemon -r -v -d $directory$newchannel &
subchild example:
-----------------
vlc atsc://frequency=605029000 --intf=dummy --sout-transcode-audio-sync :live-cache=3000 --sout='#transcode{vcodec=h264,vb=150,fps=25,width=480,scale=1,venc=x264{aud,profile=baseline,level=30,keyint=15,bframes=0,ref=1},acodec=aac,ab=40,channels=2,samplerate=22050}:duplicate{dst=std{mux=ts,dst=-,access=livehttp{seglen=16,delsegs=true,numsegs=10,index=/var/www/stream/live.m3u8,index-url=content/live-######.ts},mux=ts{use-key-frames},dst=/var/www/stream/content/live-######.ts,ratecontrol=true}}'
how can i keep the subchild from terminating??? Running Apache on Ubuntu 12.04
I got it!
For anyone interested in how, i changed my tactics to use nohup, &, disown, and > /dev/null 2>&1.
Honestly, still not quite sure how I got it working... just a lot of trial and error with some educated guesses. My code follows:
parent invocation:
------------------
nohup ./ChangeChannel.sh $channel & disown
child invocation:
-----------------
sudo nohup su user $directory$newchannel &> /dev/null 2>&1
subchild invocation:
--------------------
vlc atsc://frequency=605029000 --intf=dummy --sout-transcode-audio-sync :live-cache=3000 --sout='#transcode{vcodec=h264,vb=150,fps=25,width=480,scale=1,venc=x264{aud,profile=baseline,level=30,keyint=15,bframes=0,ref=1},acodec=aac,ab=40,channels=2,samplerate=22050}:duplicate{dst=std{mux=ts,dst=-,access=livehttp{seglen=16,delsegs=true,numsegs=10,index=/var/www/stream/live.m3u8,index-url=content/live-######.ts},mux=ts{use-key-frames},dst=/var/www/stream/content/live-######.ts,ratecontrol=true}}' & disown
ChangeChannel.sh uses sudo to execute su via cgi in order to execute vlc as user other than root. It seems a little messy but hell it works.

Any way to start Google Chrome in headless mode?

I carefully revised the list of switches at http://peter.sh/experiments/chromium-command-line-switches/#chrome-frame and I couldn't find anything that would launch Chrome in a hidden background process.
The closest I was able to is --keep-alive-for-test + custom packaged app, but the app fails to execute any passed code because (the way it reports) "no window - ChromeHidden".
TL;DR
google-chrome --headless --remote-debugging-port=9222 http://example.com
You'd also need --disable-gpu temporarily.
Tutorial:
https://developers.google.com/web/updates/2017/04/headless-chrome
There's a work in progress: https://code.google.com/p/chromium/issues/detail?id=546953
The main deliverables are:
A library which headless applications can link to to.
A sample application which demonstrates the use of headless APIs.
So it would be possible to create a simple application that runs in console without connecting to display.
Update Apr 18 '16: The work is mainly done. There's a public forum now:
https://groups.google.com/a/chromium.org/forum/#!forum/headless-dev
Documentation is being in progress:
https://chromium.googlesource.com/chromium/src/+/master/headless/README.md
Update Sep 20 '16: It looks like chrome will eventually get the "--headless" parameter:
https://bugs.chromium.org/p/chromium/issues/detail?id=612904
There was a presentation on BlinkOn 6 (June 16/17, 2016)
Update Nov 29 '16: Design doc for --headless flag: https://docs.google.com/document/d/1aIJUzQr3eougZQp90bp4mqGr5gY6hdUice8UPa-Ys90/edit#heading=h.qxqfzv2lj12s
Update Dec 13 '16: --headless flag is expected to be available in Canary builds soon
Update Mar 12 '17: Chrome 57 has a --headless flag working. Waiting for Selenium and other tools to catch up. User guide: https://chromium.googlesource.com/chromium/src/+/lkgr/headless/README.md
This guy managed to run Chrome headlessly by using Xvfb (X virtual frame buffer) to trick Chrome into thinking it was displaying a window:
http://e-method.blogspot.fr/2010/11/google-chrome-with-xvfb-headless-server.html
If you're on Linux you could try that.
So basically you need to install X virtual frame buffer and Google Chrome via:
root#localhost: ~# apt-get install xvfb imagemagick
root#localhost: ~# apt-get install google-chrome
Then run the browser on the display:
root#localhost: ~# xvfb-run --server-args='-screen 0, 1024x768x24' \
google-chrome -start-maximized http://www.example.com \
> & /dev/null &
root#localhost: ~# DISPLAY=:99 import -window root myimage.png
Or you can look at PhantomJS project which is a headless WebKit implementation.
You could set up a linux VM and use xvfb in it.
Installation on debian / ubuntu:
sudo aptitude install xvfb
Start Chrome headless and visit http://example.com :
xvfb-run --server-args='-screen 0, 1024x768x16' google-chrome
-start-maximized http://example.com > /dev/null &
Turns out it starts in headless mode if you start it as a child subprocess. Besides that:
nircmd.exe can do win hide on chrome based on its PID
Autohotkey_L can also start Chrome hidden without a taskbar button
The Chromium Embedded Framework project seems like it might fit your usecase. I don't have personal experience with the project, but I've heard good things, and it has a solid API that you should be able to exploit for your purposes.
I don't have enough reputation to comment yet, but want to let you guys know that the chrome headless mode which Vanuan mentions actually works with Selenium webdriver.
In Java you can pass the flag to chrome through chromeDriver with the following code:
ChromeOptions options = new ChromeOptions();
options.addArguments("--headless");
ChromeDriver chromeDriver = new ChromeDriver(options);
I've recently found this article which mentions several commandline options that seem to do it. Using these keywords I googled out this piece of code which seem to confirm that these options exist.
// Does not automatically open a browser window on startup (used when
// launching Chrome for the purpose of hosting background apps).
const char kNoStartupWindow[] = "no-startup-window";
// Causes Chrome to launch without opening any windows by default. Useful if
// one wishes to use Chrome as an ash server.
const char kSilentLaunch[] = "silent-launch";
I managed to successfuly run Chrome with --no-startup-window and indeed it launched without any windows. It looked like it launched properly, it spawned all typical children, but the website I tried to make it load inside didn't seem to be actually visited. It maybe possible that this headless mode is only for running apps and not for visiting sites headless*), but it looks very promising as the normal worker tree is set up, just no windows.
The second option --silent-launch made chrome process very silent. I didn't notice any children spawned and the process exited promptly. I doubt it'll be usable for this case.
After I failed my attempts with these options, I focused on less sophisticated ways. On the bottom of the list there are two options:
// Specify the initial window position: --window-position=x,y
const char kWindowPosition[] = "window-position";
// Specify the initial window size: --window-size=w,h
const char kWindowSize[] = "window-size";
I ran Chrome with options to move it completely out of the working area:
--window-size=800,600 --window-position=-800,0
and as dirty as it feels, sure it's no true headless, but still the window is out of my sight, and everything done just with chrome's startup options, without external tools sending low-level window-hide messages.
*) yes, I know try to do weird things. Essentially I tried to get rid of Chrome window that's kept by Karma during tests. I know I can switch to PhantomJS, but I specifically needed to run them in Chrome, and the window popping out was.. eh.. disturbing in the long run.
It is currently under development, you can read more information about it from here:
https://chromium.googlesource.com/chromium/src/+/master/headless/README.md
Headless Chromium is a library for running Chromium in a
headless/server environment. Expected use cases include loading web
pages, extracting metadata (e.g., the DOM) and generating bitmaps from
page contents -- using all the modern web platform features provided
by Chromium and Blink.
It currently works on Linux, there is a nice presentation.
I was also able to make chrome headless work with NightwatchJS. Here is the config that let me use it:
"chromeHeadless": {
"desiredCapabilities": {
"browserName": "chrome",
"chromeOptions": {
"args": ["--headless"],
"binary": "/Applications/Google Chrome Canary.app/Contents/MacOS/Google Chrome Canary"
}
}
}
Chrome 59 has the ability to create instance as headless.
Find the below tutorial
https://www.automation99.com/2017/07/how-to-use-chrome-headless-using.html?m=1

CMD runs EXE, launches HTML, wait until close and run final EXE

would like have BAT file that runs set-keys.EXE, launches default.html, and then when user closes html, run set-keys-back.EXE. (they are all in the same directory together). This might be run from a CD, so I might not have ability to write a flag file and then wait to see if it is deleted in order to continue. Have already tried START /WAIT but have seen that WAIT won't actually wait for GUI 32-bit applications. Have considered one batch file calling another one, still no luck. Would prefer not to have PAUSE and user have to come back to CMD just to hit a key - seems clunky. When they close out of HTML, I execute top.window.close(). would be nice if I could put some other code after that, but I think once the window is closed it's closed - no more processing. have not been able to get WShell execute to run. HTML status bar just says error on page - no info. Would love to hear your thoughts...
Update 2: I just figured out that you can launch IE directly without having to use the start command:
#echo off
rem You can use %SCRIPTDIR% to refer to the file to load, if you like
rem Note that %SCRIPTDIR% will contain a trailing slash!
set SCRIPTDIR=%~dp0
echo Testing this script...
C:\PROGRA~1\INTERN~1\iexplore.exe %SCRIPTDIR%foo.html
echo Continuing the script...
This example works for me (Windows XP 32-bit), and waits for me to close the browser window to continue.
Update: Here's an updated code block that launches Internet Explorer. Note that I use the short path to the iexplore.exe executable, and I specify the full path to the file to load:
#echo off
echo Testing this script...
start /wait /min cmd /C "C:\PROGRA~1\INTERN~1\iexplore.exe C:\foo.html"
echo Continuing the script...
Initial Answer: You mentioned trying the start /wait command, but how did you explicitly write it? The following batch script example works for me in Windows 7 x64:
#echo off
echo Testing this script...
start /wait /min cmd /C "%windir%\system32\notepad.exe foo.html"
echo Continuing the script...
In this example, the script does not continue execution until the user closes the Notepad application. The only downside here is that an extra command window pops up, but by using the /min parameter, we can start it minimized.