I have a PHP class that I use to run php script files in background, like this:
Class BackgroundScript{
public $win_path = "C:\\xampp\\htdocs\\www\\project";
public $unix_path = "/home/my_folder/project.com";
public $script = NULL;
public $command = NULL;
public $pid = NULL;
public $start_time = NULL;
public $estimated_time = NULL;
public $ellapsed_time = NULL;
public $status_file = NULL;
...
public function kill(){
$this->removeFile();
if ( self::get_os() == "windows" ){
shell_exec(" taskkill /PID ".$this->pid);
} else{
shell_exec('kill '.$this->pid);
}
}
}
That is called this way:
$argvs = " var1 var2 ";
$process = new BackgroundScript("controller.php processCSV $argvs");
Then, it creates a process that runs in background a code similar to this:
$current_process = BackgroundProcess::getByPID(getmypid());
for ($=1; $i< 100; $i++){
performLongRunningTask();
$current_process->updateStatus();
}
What I want to know is, is possible to add at the end of the loop, the following command:
$process->kill();
Considering that this php file is executed by the process I want to kill? And are there any side effects I should contemplate?
For windows you can use /F to force the kill:
taskkill /F /PID YOUR_PID
For linux you can use the syntax: kill [signal or option] PID(s) The options you can use are:
SIGHUP 1 Hangup (the default and safest way to kill a process)
SIGKILL 9 Kill Signal (less secure way of killing a process)
SIGTERM 15 Terminate (the most unsafe way to kill a process which terminates a process without saving)
When you use kill '.$this->pid it will send a SIGTERM signal. A few things can happen here:
the process may stop immediately
the process may stop after a short delay after cleaning up resources
the process may keep running indefinitely
The application can determine what it wants to do once a SIGTERM is received. While most applications will clean up their resources and stop, some may not. An application may be configured to do something completely different when a SIGTERM is received. Also, if the application is in a bad state, such as waiting for disk I/O, it may not be able to act on the signal that was sent.
Related
I am using phing and running selenium server via ExecTask. Sometimes I need to stop running server by killing its process.
Is there a possibility in phing of getting PID of process spawned in ExecTask ?
No, ExecTask cannot give the pid of spawned processes directly. It can only return it's exit status and output.
Maybe you can modify the command that you run in ExecTask itself to save the pid of spawned process. You can use $! to get the pid of the most recent background command.
job1 & //start job1 and run in background, end command with &
p1=$! //stores the pid
echo $p1 //gives pid of job1
When you want to kill the selenium server you can call this in another ExecTask :
pkill pid_to_kill
I am not sure if the changes made in shell environment with ExecTask stay or not. If yes then you can use $p1. Replace pid_to_kill with $p1 to kill job1. Else you will have to echo the pid and use the value from its output.
Otherwise you will have do pgrep name_of_program. It will give all processes containing the name. Then you can kill it with pkill.
Instead of launching the process you want to kill from the exec task (selenium server in your case). Use the exec task to launch a script (I used bash but ruby, python etc. will work too). this script will start the desired task and echo the pid. Substitute the required path and executable you want to run in the below.
#!bin/bash
./path_to_executable/process_to_run &
echo $!
Note the "&" this sends the process to the background and allows phing to continue building your project. The last line outputs the pid which can then be captured and saved to a file by the phing exec task. To save this pid add the output option to the phing exec task:
<exec command="your_script" spawn="true" output="./pid.txt" />
the output option will save the output of the exec task to a pid.txt file in the current directory. Note you may need to chown this file (to the user running phing) to enable it to be read later.
In a separate task, you can read the pid from the file and then use an exec task to kill the process.
<loadfile property="pid" file="./pid.txt" />
<exec command="kill ${pid}" dir="./" />
Note: in the above you may need to prepend sudo to the kill command (depending on who owns the process, and how it was started.
Optional but worth considering is adding a task to remove the pid.txt file. This will prevent any possibility of killing the wrong process (based on a stale pid). You may also want to sanity check the content of the pid.txt file since in the event of an error it could contain something other than the pid.
While this may not be the most direct or optimal solution it does work.
It is possible, you could go about the second parameter inside the exec command.
exec("Script To Run", $output);
The second variable gets the output of the current running script in an array format. So to show the full and readable text from the output I would use a foreach loop:
exec("ifconfig", $output); // Presuming you are developing for a Linux server
foreach ($output as $outputvar) {
echo $outputvar . "<br>";
}
After that, I would use something like strpos to pull the information from the $outputvar for the string which you are looking for.
I hope this is something similar to what you are looking for.
I ended up creating a phing task that saves the pid of the launched program and it stops it when you ask it for. It uses Cocur\BackgroundProcess to start the process in background and can also return the pid.
<?php
require_once "phing/Task.php";
class BackgroundExecTask extends Task {
protected $command = null;
protected $executable = null;
protected $id = null;
protected static $pidMap = [];
public function init() {
if (!class_exists('\Cocur\BackgroundProcess\BackgroundProcess')) {
throw new BuildException("This task requires the Cocur Background Process componente installed and available on the include path", $this->getLocation());
}
}
public function main() {
switch ($this->command) {
case "start":
return $this->start();
case "stop":
return $this->stop();
}
}
protected function start() {
$process = new \Cocur\BackgroundProcess\BackgroundProcess($this->executable);
$process->run();
// you can also return the pid
//$this->project->setProperty($this->pidProperty, $process->getPid());
self::$pidMap[$this->id] = $process;
}
protected function stop() {
self::$pidMap[$this->id]->stop();
}
public function setCommand($command)
{
$this->command = "" . $command;
}
public function setExecutable($executable)
{
$this->executable = "" . $executable;
}
public function setId($id)
{
$this->id = "" . $id;
}
}
Usage:
<backgroundexec id="myprogram" command="start" executable="somebinary" />
<backgroundexec id="myprogram" command="stop" />
I'm opening a long-running process with popen(). For debugging, I'd like to terminate the process before it has completed. Calling pclose() just blocks until the child completes.
How can I kill the process? I don't see any easy way to get the pid out of the resource that popen() returns so that I can send it a signal.
I suppose I could do something kludgey and try to fudge the pid into the output using some sort of command-line hackery...
Well, landed on a solution: I switched back to proc_open() instead of popen(). Then it's as simple as:
$s = proc_get_status($p);
posix_kill($s['pid'], SIGKILL);
proc_close($p);
Just send a kill (or abort) signal using kill function:
php http://php.net/manual/en/function.posix-kill.php
c/c++ http://linux.die.net/man/3/kill
You can find the pid, and checks that you're really its parent by doing:
// Find child processes according to current pid
$res = trim(exec('ps -eo pid,ppid |grep "'.getmypid().'" |head -n2 |tail -n1'));
if (preg_match('~^(\d+)\s+(\d+)$~', $res, $pid) !== 0 && (int) $pid[2] === getmypid())
{
// I'm the parent PID, just send a KILL
posix_kill((int) $pid[1], 9);
}
It's working quite well on a fast-cgi PHP server.
I have a cron script that executes a PHP script every 10 minutes. The script checks a queue and processes the data in the queue. Sometimes the queue has enough data to last over 10 minutes of processing, creating the potential of two scripts trying to access the same data. I want to be able to detect whether the script is already running to prevent launching multiple copies of the script. I thought about creating a database flag that says that a script is processing, but if the script were ever to crash it would leave it in the positive state. Is there an easy way to tell if the PHP script is already running from withing a PHP or shell script?
You can just use a lock file. PHP's flock() function provides a simple wrapper for Unix's flock function, which provides advisory locks on files.
If you don't explicitly release them, the OS will automatically release these locks for you when the process holding them terminates, even if it terminates abnormally.
You can also follow the loose Unix convention of making your lock file a 'PID file' - that is, upon obtaining a lock on the file, have your script write its PID to it. Even if you never read this from within your script, it will be convenient for you if your script ever hangs or goes crazy and you want to find its PID in order to manually kill it.
Here's a copy/paste-ready implementation:
#!/usr/bin/php
<?php
$lock_file = fopen('path/to/yourlock.pid', 'c');
$got_lock = flock($lock_file, LOCK_EX | LOCK_NB, $wouldblock);
if ($lock_file === false || (!$got_lock && !$wouldblock)) {
throw new Exception(
"Unexpected error opening or locking lock file. Perhaps you " .
"don't have permission to write to the lock file or its " .
"containing directory?"
);
}
else if (!$got_lock && $wouldblock) {
exit("Another instance is already running; terminating.\n");
}
// Lock acquired; let's write our PID to the lock file for the convenience
// of humans who may wish to terminate the script.
ftruncate($lock_file, 0);
fwrite($lock_file, getmypid() . "\n");
/*
The main body of your script goes here.
*/
echo "Hello, world!";
// All done; we blank the PID file and explicitly release the lock
// (although this should be unnecessary) before terminating.
ftruncate($lock_file, 0);
flock($lock_file, LOCK_UN);
Just set the path of your lock file to wherever you like and you're set.
If you need it to be absolutely crash-proof, you should use semaphores, which are released automatically when php ends the specific request handling.
A simpler approach would be to create a DB record or a file at the beginning of the execution, and remove it at the end. You could always check the "age" of that record/file, and if it's older than say 3 times the normal script execution, suppose it crashed and remove it.
There's no "silver bullet", it just depends on your needs.
If you are running Linux, this should work at the top of your script:
$running = exec("ps aux|grep ". basename(__FILE__) ."|grep -v grep|wc -l");
if($running > 1) {
exit;
}
A common way for *nix daemons (though not necessarily PHP scripts, but it will work) is to use a .pid file.
When the script starts check for the existence of a .pid file named for the script (generally stored in /var/run/). If it doesn't exist, create it setting its contents to the PID of the process running the script (using getmypid) then continue with normal execution. If it does exist read the PID from it and see if that process is still running, probably by running ps $pid. If it is running, exit. Otherwise, overwrite its contents with your PID (as above) and continue normal execution.
When execution finished, delete the file.
I know this is an old question but in case someone else is looking here I'll post some code. This is what I have done recently in a similar situation and it works well. Put put this code at the top of your file and if the same script is already running it will leave it be and end the new one.
I use it to keep a monitoring system running at all times. A cron job starts the script every 5 minutes but unless the other has stopped from some reason (usually if it has crashed, which is very rare!) the new one will just exit itself.
// The file to store our process file
define('PROCESS_FILE', 'process.pid');
// Check I am running from the command line
if (PHP_SAPI != 'cli') {
log_message('Run me from the command line');
exit;
}
// Check if I'm already running and kill myself off if I am
$pid_running = false;
if (file_exists(PROCESS_FILE)) {
$data = file(PROCESS_FILE);
foreach ($data as $pid) {
$pid = (int)$pid;
if ($pid > 0 && file_exists('/proc/' . $pid)) {
$pid_running = $pid;
break;
}
}
}
if ($pid_running && $pid_running != getmypid()) {
if (file_exists(PROCESS_FILE)) {
file_put_contents(PROCESS_FILE, $pid);
}
log_message('I am already running as pid ' . $pid . ' so stopping now');
exit;
} else {
// Make sure file has just me in it
file_put_contents(PROCESS_FILE, getmypid());
log_message('Written pid with id '.getmypid());
}
It will NOT work without modification on Windows, but should be fine in unix based systems.
You can use new Symfony 2.6 LockHandler.
Source
$lock = new LockHandler('update:contents');
if (!$lock->lock()) {
echo 'The command is already running in another process.';
}
This worked for me. Set a database record with a lock flag and a time stamp. My script should complete well within 15min so added that as a last locked feild to check:
$lockresult = mysql_query("
SELECT *
FROM queue_locks
WHERE `lastlocked` > DATE_SUB(NOW() , INTERVAL 15 MINUTE)
AND `locked` = 'yes'
AND `queid` = '1'
LIMIT 1
");
$LockedRowCount = mysql_num_rows($lockresult);
if($LockedRowCount>0){
echo "this script is locked, try again later";
exit;
}else{
//Set the DB record to locked and carry on son
$result = mysql_query("
UPDATE `queue_locks` SET `locked` = 'yes', `lastlocked` = CURRENT_TIMESTAMP WHERE `queid` = 1;
");
}
Then unlock it at the end of the script:
$result = mysql_query("UPDATE `queue_locks` SET `locked` = 'no' WHERE `queid` = 1;");
I know this is an old question, but there's an approach which hasn't been mentioned before that I think is worth considering.
One of the problems with a lockfile or database flag solution, as already mentioned, is that if the script fails for some reason other than normal completion it won't release the lock. And therefore the next instance won't start until the lock is either manually cleared or cleared by a clean-up function.
If, though, you are certain that the script should only ever be running once, then it's relatively easy to check from within the script whether it is already running when you start it. Here's some code:
function checkrun() {
exec("ps auxww",$ps);
$r = 0;
foreach ($ps as $p) {
if (strpos($p,basename(__FILE__))) {
$r++;
if ($r > 1) {
echo "too many instances, exiting\n";
exit();
}
}
}
}
Simply call this function at the start of the script, before you do anything else (such as open a database handler or process an import file), and if the same script is already running then it will appear twice in the process list - once for the previous instance, and once for this one. So, if it appears more than once, just exit.
A potential gotcha here: I'm assuming that you will never have two scripts with the same basename that may legitimately run simultaneously (eg, the same script running under two different users). If that is a possibility, then you'd need to extend the checking to something more sophisticated than a simple substring on the file's basename. But this works well enough if you have unique filenames for your scripts.
Assuming this is a linux server and you have cronjobs available
///Check for running script and run if non-exist///
#! /bin/bash
check=$(ps -fea | grep -v grep | grep script.php | wc -l)
date=$(date +%Y-%m%d" "%H:%M:%S)
if [ "$check" -lt 1 ]; then
echo "["$date"] Starting script" >> /path/to/script/log/
/sbin/script ///Call the script here - see below///
fi
script file
#/usr/bin/php /path/to/your/php/script.php
Home / Check if a PHP script is already running
Check if a PHP script is already running
If you have long running batch processes with PHP that are run by cron and you want to ensure there’s only ever one running copy of the script, you can use the functions getmypid() and posix_kill() to check to see if you already have a copy of the process running. This post has a PHP class for checking if the script is already running.
Each process running on a Linux/Unix computer has a pid, or process identifier. In PHP this can be retrieved using getmypid() which will return an integer number. This pid number can be saved to a file and each time the script is run a check made to see if the file exists. If it is the posix_kill() function can be used to see if a process is running with that pid number.
My PHP class for doing this is below. Please feel free to use this and modify to suit your individual requirements.
class pid {
protected $filename;
public $already_running = false;
function __construct($directory) {
$this->filename = $directory . '/' . basename($_SERVER['PHP_SELF']) . '.pid';
if(is_writable($this->filename) || is_writable($directory)) {
if(file_exists($this->filename)) {
$pid = (int)trim(file_get_contents($this->filename));
if(posix_kill($pid, 0)) {
$this->already_running = true;
}
}
}
else {
die("Cannot write to pid file '$this->filename'. Program execution halted.n");
}
if(!$this->already_running) {
$pid = getmypid();
file_put_contents($this->filename, $pid);
}
}
public function __destruct() {
if(!$this->already_running && file_exists($this->filename) && is_writeable($this->filename)) {
unlink($this->filename);
}
}
}
Use Class below
$pid = new pid('/tmp');
if($pid->already_running) {
echo "Already running.n";
exit;
}
else {
echo "Running...n";
}
Inspired by Mark Amery's answer I created this class. This might help someone. Simply change the "temp/lockFile.pid" to where you want the file placed.
class ProcessLocker
{
private $lockFile;
private $gotLock;
private $wouldBlock;
function __construct()
{
$this->lockFile = fopen('temp/lockFile.pid', 'c');
if ($this->lockFile === false) {
throw new Exception("Unable to open the file.");
}
$this->gotLock = flock($this->lockFile, LOCK_EX | LOCK_NB, $this->wouldBlock);
}
function __destruct()
{
$this->unlockProcess();
}
public function isLocked()
{
if (!$this->gotLock && $this->wouldBlock) {
return true;
}
return false;
}
public function lockProcess()
{
if (!$this->gotLock && !$this->wouldBlock) {
throw new Exception("Unable to lock the file.");
}
ftruncate($this->lockFile, 0);
fwrite($this->lockFile, getmypid() . "\n");
}
public function unlockProcess()
{
ftruncate($this->lockFile, 0);
flock($this->lockFile, LOCK_UN);
}
}
Simply use the class as such in the beginning of your script:
$locker = new ProcessLocker();
if(!$locker->isLocked()){
$locker->lockProcess();
} else{
// The process is locked
exit();
}
I've got a PHP script in the works that is a job worker; its main task is to check a database table for new jobs, and if there are any, to act on them. But jobs will be coming in in bursts, with long gaps in between, so I devised a sleep cycle like:
while(true) {
if ($jobs = get_new_jobs()) {
// Act upon the jobs
} else {
// No new jobs now
sleep(30);
}
}
Good, but in some cases that means there might be a 30 second lag before a new job is acted upon. Since this is a daemon script, I figured I'd try the pcntl_signal hook to catch a SIGUSR1 signal to nudge the script to wake up, like:
$_isAwake = true;
function user_sig($signo) {
global $_isAwake;
daemon_log("Caught SIGUSR1");
$_isAwake = true;
}
pcntl_signal(SIGUSR1, 'user_sig');
while(true) {
if ($jobs = get_new_jobs()) {
// Act upon the jobs
} else {
// No new jobs now
daemon_log("No new jobs, sleeping...");
$_isAwake = false;
$ts = time();
while(time() < $ts+30) {
sleep(1);
if ($_isAwake) break; // Did a signal happen while we were sleeping? If so, stop sleeping
}
$_isAwake = true;
}
}
I broke the sleep(30) up into smaller sleep bits, in case a signal doesn't interrupt a sleep() command, thinking that this would cause at most a one-second delay, but in the log file, I'm seeing that the SIGUSR1 isn't being caught until after the full 30 seconds has passed (and maybe the outer while loop resets).
I found the pcntl_signal_dispatch command, but that's only for PHP 5.3 and higher. If I were using that version, I could stick a call to that command before the if ($_isAwake) call, but as it currently stands I'm on 5.2.13.
On what sort of situations is the signals queue interpreted in PHP versions without the means to explicitly call the queue parsing? Could I put in some other useless command in that sleep loop that would trigger a signal queue parse within there?
Fixed my own problem: The answer is the "ticks" declaration. I had, as part of the Daemon process startup done the declare(ticks=1); action, but it wasn't seeming to carry over to the main script (since that was inside a function, in an include file?. Adding a declare(ticks=1) line before the while(true) loop causes signals to come through immediately (i.e. the sleep(1) command causes a tick, so after waking up from sleep, signals are processed).
I am converting a PDF with PDF2SWF and Indexing with XPDF.. with exec.. only this requires the execution time to be really high.
Is it possible to run it as background process and then launch a script when it is done converting?
in general, php does not implement threads.
But there is an ZF-class which may be suitable for you:
http://framework.zend.com/manual/en/zendx.console.process.unix.overview.html
ZendX_Console_Process_Unix allows
developers to spawn an object as a new
process, and so do multiple tasks in
parallel on console environments.
Through its specific nature, it is
only working on nix based systems
like Linux, Solaris, Mac/OSx and such.
Additionally, the shmop_, pcntl_* and
posix_* modules are required for this
component to run. If one of the
requirements is not met, it will throw
an exception after instantiating the
component.
suitable example:
class MyProcess extends ZendX_Console_Process_Unix
{
protected function _run()
{
// doing pdf and flash stuff
}
}
$process1 = new MyProcess();
$process1->start();
while ($process1->isRunning()) {
sleep(1);
}
echo 'Process completed';
.
Try using popen() instead of exec().
This hack will work on any standard PHP installation, even on Windows, no additional libraries required. Yo can't really control all aspects of the processes you spawn this way, but sometimes this is enough:
$p1 = popen("/bin/bash ./some_shell_script.sh argument_1","r");
$p2 = popen("/bin/bash ./some_other_shell_script.sh argument_2","r");
$p2 = popen("/bin/bash ./yet_other_shell_script.sh argument_3","r");
The three spawned shell scripts will run simultaneously, and as long as you don't do a pclose($p1) (or $p2 or $p3) or try to read from any of these pipes, they will not block your PHP execution.
When you're done with your other stuff (the one that you are doing with your PHP script) you can call pclose() on the pipes, and that will pause your script execution until the process you are pclosing finishes. Then your script can do something else.
Note that your PHP will not conclude or die() until those scripts have finished. Reaching the end of the script or calling die() will make it wait.
If you are running it from the command line, you can fork a php process using pcntl_fork
There are also daemon classes that would do the same trick:
http://pear.php.net/package/System_Daemon
$pid = pcntl_fork();
if ($pid == -1) {
die('could not fork');
} else if ($pid) {
//We are the parent, exit
exit();
} else {
// We are the child, do something interesting then call the script at the end.
}