PHP testing between pthreads and curl - php

Question

We are planning to building real time bidding and we are evaluating performance of PHP compare to Java in terms of throughput/response times etc.
(Java part is taken care by other member of team)
Initial start:
I have a test script which makes 50 http connection to different servers.
1st approach
- I am using curl_multi_init function and I get response under 7 seconds.
2nd approach
- I am using PHP pthreads api and trying to make parallel calls and expecting same response time or less.But total time on average is around 25 seconds
Here is the code
<?php
$g_request_arr = array(
'0' => array(
'request_url' => 'https://www.google.co.uk/?#q=56%2B12'
),
..
..
..
'49'=>array(
'request_url' => 'https://www.google.co.uk/?#q=256%2B132'
)
);
class ChildThread extends Thread {
public function __construct($urls) {
$this->data = $urls;
}
public function run(){
foreach($this->data as $url_info ){
$url = $url_info['request_url'];
file_get_contents($url);
}
$this->synchronized(function($thread){
$thread->notify();
}, $this);
}
}
$thread = new ChildThread($g_request_arr);
$thread->start();
$thread->synchronized(function($thread){
}, $thread);
?>
I want to know what is missing in above code or is it possible to bring the response under 7 seconds.

Answer 1

You are requesting all the data in one thread, here's a better approach:
<?php
class WebRequest extends Stackable {
public $request_url;
public $response_body;
public function __construct($request_url) {
$this->request_url = $request_url;
}
public function run(){
$this->response_body = file_get_contents(
$this->request_url);
}
}
class WebWorker extends Worker {
public function run(){}
}
$list = array(
new WebRequest("http://google.com"),
new WebRequest("http://www.php.net")
);
$max = 8;
$threads = array();
$start = microtime(true);
/* start some workers */
while (#$thread++<$max) {
$threads[$thread] = new WebWorker();
$threads[$thread]->start();
}
/* stack the jobs onto workers */
foreach ($list as $job) {
$threads[array_rand($threads)]->stack(
$job);
}
/* wait for completion */
foreach ($threads as $thread) {
$thread->shutdown();
}
$time = microtime(true) - $start;
/* tell you all about it */
printf("Fetched %d responses in %.3f seconds\n", count($list), $time);
$length = 0;
foreach ($list as $listed) {
$length += strlen($listed["response_body"]);
}
printf("Total of %d bytes\n", $length);
?>
This uses multiple workers, which you can adjust by changing $max. There's not much point in creating 1000 threads if you have 1000 requests to process.