I want to run service in PHP which to be run in the background. I have tried by using exec() function in PHP but service is being run in infinite loop and control is not returning back over PHP file. I have searched more over the internet but I can't find the solution. Please give me some idea or reference to achieve this task.
This is code for reference what I want to do:-
echo"hello";
exec("raintree.frm");
echo"hello1";
raintree.frm is a service which I want to execute. Here PHP script prints "hello" over browser but that is not coming on "hello1" because control gets stuck on exec() function.
If you'd like to have your service running in a separate process, as the title states, you need to create the new process and then run the service in it. In PHP you can create a new process with pcntl_fork() and start the service in the child process. Something like this
echo "hello";
$pid = pcntl_fork();
switch($pid){
case -1: // pcntl_fork() failed
die('could not fork');
case 0: // you're in the new (child) process
exec("raintree.frm");
// controll goes further down ONLY if exec() fails
echo 'exec() failed';
default: // you're in the main (parent) process in which the script is running
echo "hello1";
}
For more clarification read the manual (the link above to pcntl_fork()) as well as look at some C/Unix tutorials on the topics (or rather syscalls) fork() and exec().
You can make use of atd for running arbitrary commands in a separate process:
shell_exec('echo code you want to run | at -m now');
For this, atd should be installed and running, of course.
The exec() function is waiting to receive the output of the externally executed command.
To prevent this from happening, you can redirect the output elsewhere:
<?php
echo 'hello1';
exec('raintree.frm > /dev/null &');
echo 'hello2';
This example will output "hello1hello2" without waiting for raintree.frm > /dev/null & to finish.
(this probably only works with Unix-like operating systems)
Related
I need to echo text to a named pipe (FIFO) in Linux. Even though I'm running in background with '&' and redirecting all output to a /dev/null, the shell_exec call always blocks.
There are tons of answers to pretty much exactly this question all over the internet, and they all basically point to the following php manual section:
If a program is started with this function, in order for it to continue running in the background, the output of the program must be redirected to a file or another output stream. Failing to do so will cause PHP to hang until the execution of the program ends.
And sure enough, when I try the non-blocking approach (of backgrounding and redirecting to /dev/null) with other commands like sleep, php successfully executes without hanging. But for the case of echo-ing to the FIFO, php hangs even though running the same command with bash produces no visible output and immediately returns to the shell.
In bash, I can run:
bash$ { echo yay > fifo & } &> /dev/null
bash$ cat fifo
yay
[1]+ Done echo yay > fifo
but when running the following php file with php echo.php:
<?php
shell_exec("{ echo yay > fifo & } &> /dev/null");
?>
it hangs, unless I first open fifo for reading.
So my question is, why is this blocking, but sleep isn't? In addition, I want to know what is happening behind the scenes: when I put the '&' in the php call, even though the shell_exec call blocks, the echo call clearly doesn't block whatever bash session php invoked it on, because when I CTRL+C out of php, I can read 'yay' from the FIFO (if I don't background the echo command, after CTRL+C the FIFO contains no text). This suggests that perhaps php is waiting on the pid of the echo command before going to the next instruction. Is this true?
I've been trying something similar and in the end came up with this solution:
/**
* This runs a shell command on the server under the current PHP user, that is in CLI mode it is the user you are logged in with.
* If a command is run in the background the method will return the location of the tempfile that captures the output. In that case you will have to manually remove the temporary file.
*/
static public function command($cmd, $show_output = true, $escape_command = false, $run_in_background = false)
{
if ($escape_command)
$cmd = escapeshellcmd($cmd);
$f = trim(`mktemp`);
passthru($cmd . ($show_output ? " | tee $f" : " > $f") . ($run_in_background ? ' &' : ''));
return $run_in_background ? $f : trim(`cat $f ; rm -rf $f`);
}
The trick is to write the output to a temporary file and return that when the command has finished (blocking behavior) or just return the file path (non-blocking behavior). Also, I'm using passthru rather than shell_exec because interactive sessions are not possible with the latter because of the blocking behavior.
I am currently following a tutorial that teaches how to create a queue in php. An infinite loop was created in a php script. I simplified the code in order to focus on the question at hand:
while(1) {
echo 'no jobs to do - waiting...', PHP_EOL;
sleep(10);
}
I use PuTTy (with an SSH connection) to connect to the linux terminal in my shared hosting account (godaddy). If I run php queuefile.php, I know it will run with no problems (already tested the code with a finite for loop instead of the infinite while loop).
QUESTION: How could I exit out of the infinite loop once it has started? I have already read online the option of creating code that "checks" if it should continue looping with something like the following code:
$bool = TRUE;
while ($bool)
{
if(!file_exists(allow.txt)){$bool = FALSE}
//... the rest of the code
though I am curious if there might be a command I can type in the terminal, or a set of keys I can push that will cause the script to terminate. If there is any way of terminating the script, or if there is a better way to make the previous "check", I would love your feedback!
Pushing Ctrl+C should stop the running program that is running in the foreground.
You could also kill it when you login in another session an do some ps aux | grep my-php-script.php and see if it is your program, then you can use pkill -f my-php-script.php to kill this process.
I understand so that you want to make cron in your server. Therefore you should log in your server via putty and create cron job.
For example:
After logging...
crontab -e
Then add
1 2 3 4 5 /path/to/command arg1 arg2
I am wanting to execute a large, database intensive script, but do not need to wait for the process to finish. I would simply like to call the script, let it run in the background and then redirect to another page.
EDIT:
i am working on a local Zend community server, on Windows 7.
I have access to remote linux servers where the project also resides, so i can do this on linux or windows.
i have this
public function createInstanceAction()
{
//calls a seperate php process which creates the instance
exec('php -f /path/to/file/createInstance.php');
Mage::getSingleton('adminhtml/session')->addSuccess(Mage::helper('adminhtml')->__('Instance creation process started. This may take up to a few minutes.'));
$this->_redirect('instances/adminhtml_instances/');
return;
}
this works perfectly, but the magento application hangs around for the process to finish. it does everything i expect, logging to file from time to time, and am happy with how its running. Now all i would like to do is have this script start, the controller action does not hang around, but instead redirects and thats that. from what I have learnt about exec(), you can do so by changing the way i call exec() above, to :
exec('php -f /path/to/file/createInstance.php > /dev/null 2>&1 &');
which i took from here
if i add "> /dev/null 2>&1 &" to the exec call, it doesnt wait around as expected, but it does not execute the script anymore. Could someone tell me why, and if so, tell me how i can get this to work please?
Could this be a permission related issue?
thanks
EDIT : Im assuming it would be an issue to have any output logged to file if i call the exec function with (/dev/null 2>&1 &) as that would cancel that. is that correct?
After taking time to fully understand my own question and the way it could be answered, i have prepared my solution.
Thanks to all for your suggestions, and for excusing my casual, unpreparedness when asking the question.
The answer to the above question depends on a number of things, such as the operating system you are referring to, which php modules you are running and even as far as what webserver you are running. So if i had to start the question again, the first thing i would do is state what my setup is.
I wanted to achieve this on two environments :
1.) Windows 7 running Zend server community edition.
2.) Linux (my OS is Linux odysseus 2.6.32-5-xen-amd64 #1 SMP Fri Sep 9 22:23:19 UTC 2011 x86_64)
to get this right, i wanted it to work either way when deploying to windows or linux, so i used php to determine what the operating system was.
public function createInstanceAction()
{
//determines what operating system is Being used
if (strtoupper(substr(PHP_OS, 0, 3)) === 'WIN')
{
//This is a windows server
//call a seperate php process to run independently from the broswer action
pclose(popen("start php /path/to/script/script.php","r"));
}
else
{
//assuming its linux, but in fact it simply means its not windows
// to check for linux specifically use (strtoupper(substr(PHP_OS, 0, 3)) === 'LIN')
exec('php -f /path/to/file/script.php >/dev/null 2>&1 &');
}
//the browser will not hang around for this process to complete, and you can contimue with whatever actions you want.
//myscript log any out put so i can capture info as it runs
}
In short, ask questions once you understand them. there are many ways to to achieve the above, and this is just one solution that works for my development and production environments.
thanks for the help all.
PHP popen
From the docs (this should help you do other stuff, while that process is working; not sure if closing the current PHP process will kill the opened process):
/* Add redirection so we can get stderr. */
$handle = popen('/path/to/executable 2>&1', 'r');
echo "'$handle'; " . gettype($handle) . "\n";
$read = fread($handle, 2096);
echo $read;
pclose($handle);
Solution 2:
Trick the browser to close the connection (assuming there is a browser involved):
ob_start();
?><html><!--example html body--></html><?php
$strContents=ob_get_clean();
header("Connection: Close");
header("Content-encoding: none");//doesn't work without this, I don't know why:(
ignore_user_abort(true);
header("Content-type: text/html");
header("Content-Length: ".strlen($strContents));
echo $strContents;
flush();
//at this point a real browser would close the connection and finish rendering;
//crappy http clients like some curl implementations (and not only) would wait for the server to close the connection, then finish rendering/serving results...:(
//TODO: add long running operations here, exec, or whatever you have.
You could write a wrapper-script, say createInstance.sh like
#! /bin/bash
trap "" SIGHUP
php -f "$1" > logfile.txt 2>&1 &
Then you call the script from within PHP:
exec('bash "/path/to/file/createInstance.sh"');
which should detach the new php process most instantly from the script. If that doesen't help, you might try to use SIGABRT, SIGTERM or SIGINT instead of SIGHUP, I don't know exactly which signal is sent.
I've been able to use:
shell_exec("nohup $command > /dev/null & echo $!")
Where $command is for example:
php script.php --parameter 1
I've noticed some strange behavior with this. For example running mysql command line doesn't work, only php scripts seem to work.
Also, running cd /path/to/dir && php nohup $command ... doesn't work either, I had to chdir() within the PHP script and then run the command for it to work.
The PHP executable included with Zend Server seems to be what's causing attempts to run a script in the background (using the ampersand & operator in the exec) to fail.
We tested this using our standard PHP executable and it worked fine. It's something to do with the version shipped with Zend Server though our limited attempts to figure out what that was going on have not turned anything up.
This question already has answers here:
Closed 11 years ago.
Possible Duplicate:
Asynchronous shell exec in PHP
i need to run a java program in the background.
process.php contains
shell_exec("php php_cli.php")
php_cli.php contains
shell_exec("java -jar BiForce.jar settings.ini > log.txt");
I am calling process.php asynchronously using ajax
When i click the link in the webpage that calls ajax function (for running process.php) the webage shows "loading". when i click other links at the same time it does not responds.
The java program takes about 24 hours to finish executing, so user will not wait until the execution ends.
The problem is that the browser keeps on loading and does not go to other pages when clicked the link.
I also tried with system(), but the same problem ....
Help will greatly be appreciated.
Using shell_exec waits for the command to hang up, so that's what your script is doing.
If your command doesn't have any wait time, then your script will not either.
You can call another PHP script from your original, without waiting for it to hang up:
$processId = shell_exec(
"nohup " . // Runs a command, ignoring hangup signals.
"nice " . // "Adjusted niceness" :) Read nice --help
"/usr/bin/php -c " . // Path to your PHP executable.
"/path/to/php.ini -f " . // Path to your PHP config.
"/var/www/php_cli.php " . // Path to the script you want to execute.
"action=generate > /process.log " . // Log file.
"& echo $!" // Make sure it returns only the process id.
);
It is then possible to detect whether or not the script is finished by using this command:
exec('ps ' . $processId, $processState);
// exec returns the result of the command - but we need to store the process state.
// The third param is a referenced variable.
// First key in $processState is that it's running.
// Second key would be that it has exited.
if (count($processState) < 2) {
// Process has ended.
}
You could call the command in the page displayed, but appending an & at the end:
shell_exec("java -jar BiForce.jar settings.ini > log.txt &");
This way the process is launched on the background.
Also, there is no need (unless defined by your application) to create a process.php wich itself calls php via a shell exec. You could archive the same functionality via an include to the other file.
As in normal shell scripting you can use the ampersand to background the process:
shell_exec("java -jar BiForce.jar settings.ini > log.txt &");
See Asynchronous shell exec in PHP .
First, you might want to redesign this concept. I am not sure exactly what these programs do, but clearly this is can lead to potential problems...
This is what I suggest you do, instead of starting external processes via PHP:
Your ajax call creates (or reuse) a file in some temporary directory (probably using the user session to generate that file)
some data is written unto the file, and the request ends
Your jar is launched separately, and runs indefinitely
At regular intervals, the Java program scans the temporary directory for new files, or if some file has been modified
parse it, and execute the 24 hour long process, or adjust any previous execution if necessary
Along the same idea, you can even use sockets instead to communicate with that Java program, or any other way.
The advantage of having the Java program running all the time instead of starting a new process is to be able to reuse system resources within the lifetime of the application; for example, if your program is using DB connections, or any data, cache, etc.
I am wanting to use ffmpeg to convert video to .flv in php. Currently I have this working, but it hangs the browser until the file is uploaded and is finished. I have been looking at the php docs on how to run an exec() process in the background, while updating the process using the returned PID. Here is what I found:
//Run linux command in background and return the PID created by the OS
function run_in_background($Command, $Priority = 0)
{
if($Priority)
$PID = shell_exec("nohup nice -n $Priority $Command > /dev/null & echo $!");
else
$PID = shell_exec("nohup $Command > /dev/null & echo $!");
return($PID);
}
There is also a trick which I use to track if the background task is running using the returned PID :
//Verifies if a process is running in linux
function is_process_running($PID)
{
exec("ps $PID", $ProcessState);
return(count($ProcessState) >= 2);
}
Am I suppose to create a separate .php file which then runs from the php cli to execute one of these functions? I just need a little nudge in getting this working and then I can take it from there.
Thanks!
Am I suppose to create a separate .php
file which then runs from the php cli
to execute one of these functions?
This is probably the way I would do it :
the PHP webpage adds a record in database to indicate "this file has to be processed"
and displays a message to the user ; something like "your file will be processed soon"
In CLI, have a batch process the new inserted files
first, mark a record as "processing"
do the ffmpeg thing
mark the file as "processed"
And, on the webpage, you can show to the user in which state his file is :
if it has not been processed yet
if it's being processed
or if it's been processed -- you can then give him the link to the new video file.
Here's a couple of other thoughts :
The day your application becomes bigger, you can have :
one "web server"
many "processing servers" ; in your application, it's the ffmpeg thing that will require lots of CPU, not serving web pages ; so, being able to scale that part is nice (that's another to "lock" files, indicating them as "processing" in DB : that way, you will not have several processing servers trying to process the same file)
You only use PHP from the web server to generate web pages, which is je job of a web server
Heavy / long processing is not the job of a web server !
The day you'll want to switch to something else than PHP for the "processing" part, it'll be easier.
Your "processing script" would have to be launch every couple of minutes ; you can use cron for that, if you are on a Linux-like machine.
Edit : a bit more informations, after seeing the comment
As the processing part is done from CLI, and not from Apache, you don't need anykind of "background" manipulations : you can just use shell_exec, which will return the whole ouput of the command to your PHP script when it's finished doing it's job.
For the user watching the web page saying "processing", it will seem like background processing ; and, in a way, it'll be, as the processing will be done by another processus (maybe even on another machine).
But, for you, it'll be much simpler :
one webpage (nothing "background")
one CLI script, with no background stuff either.
Your processing script could look like something like this, I suppose :
// Fetch informations from DB about one file to process
// and mark it as "processing"
// Those would be fetched / determined from the data you just fetched from DB
$in_file = 'in-file.avi';
$out_file = 'out-file.avi';
// Launch the ffmpeg processing command (will probably require more options ^^ )
// The PHP script will wait until it's finished :
// No background work
// No need for any kind of polling
$output = shell_exec('ffmpeg ' . escapeshellarg($in_file) . ' ' . escapeshellarg($out_file));
// File has been processed
// Store the "output name" to DB
// Mark the record in DB as "processed"
Really easier than what you first thought, isn't it ? ;-)
Just don't worry about the background stuff anymore : only thing important is that the processing script is launched regularly, from crontab.
Hope this helps :-)
You don't need to write a separate php script to do this (Though you may want to later if you implement some sort of queuing system).
You're almost there. The only problem is, the shell_exec() call blocks to wait for the return of the shell. You can avoid this if you redirect all output from the command in the shell to wither a file or /dev/null and background the task (with the & operator).
So your code would become:
//Run linux command in background and return the PID created by the OS
function run_in_background($Command, $Priority = 0)
{
if($Priority) {
shell_exec("nohup nice -n $Priority $Command 2> /dev/null > /dev/null &");
} else {
shell_exec("nohup $Command 2> /dev/null > /dev/null &");
}
}
I don't think there is any way to retrieve the PID, unfortunately.