I had a php command line process spawning child processes and [...]]]> The SSL/TLS process is a heavy one, it involves algorithm negotiation between client and server, key exchanges, cyphering, decyphering and authentication. But what’s surprising is, that the server you’re connecting to can directly influence the performance of your client and its CPU consumption.

I had a php command line process spawning child processes and connecting through SSL to a web server, in 2 scenarios. The first scenario was to an out of the box Apache httpd server with mod_ssl, and the second scenario was to an out of the box Nginx with the SSL module. Both were using the exact same box, and were “out of the box” meaning I used the default configuration for both.

In the first scenario I was able to spawn no more than 6 (!) php processes before the box running them began to show load, and the CPU queue started to fill up. Each php child was taking between 15%-30% cpu at any given moment.

In the second scenario, I was able to spawn 40 (!!) php child processes without the box being loaded. Each php child was taking around 1.5% cpu.

I’m no SSL expert, and there might be a way to configure Apache to inflict less load on the connecting client. There is also SSLSessionCache which might relieve load from both the server and the client. But the “out of the box” configuration shows that Nginx is a real winner again.

If you can, avoid SSL altogether. If not, terminate it at a front-end before proceeding to Apache.

It is important to note that the code on his demo uses window.location.hash, although it seems this practice [...]]]> I set out to write a short tutorial about cross domain ajax calls and communication between iframes using url hashes (fragments), but I found this great tutorial by Michael Mahemoff covering the subject inside and out.

It is important to note that the code on his demo uses window.location.hash, although it seems this practice should currently be avoided because of a known open bugs in the way Firefox URI-decodes document.location.hash when getting its value (could be considered an expected behavior and not a bug, but still different than any other browser). Facebook’s code uses substring on pathname, and it seems to work cross browser. Facebook also have on their wiki a nice visual explanation of how it works:

Until the HTML 5 postMessage is widely adopted, we will have to keep using this somewhat of a hack, but it works, that’s what’s important.

Here is a problem: There is a batch job reading remote data through XML-RPC, and updating a local MySQL database according to the XML-RPC responses. The db is InnoDB, and the entire batch job should be [...]]]> The following concept is still under examination, but my initial tests proved successful, so I thought it’s time to share.

Here is a problem: There is a batch job reading remote data through XML-RPC, and updating a local MySQL database according to the XML-RPC responses. The db is InnoDB, and the entire batch job should be transacted. That is, in any case of failure, there should be rollback, and on success there should be commit.

So, the simple way is of course a single process script that uses a linear workflow:

1. START TRANSACTION locally.
2. Make XML-RPC request, fetch data.
3. Insert data into db as needed.
4. Repeat 2-3 until Error or Finish.
5. If Error ROLLBACK, if Finish COMMIT.

This works, but you may notice a bottleneck, being the XML-RPC request. It’s using http, and it’s connecting to a remote server. Sometimes the XML-RPC server also takes time to perform the work that generates the response. Add the network latency, and you get a single process that most of the time sits idle and waits for response.

So if we have a process that just sits and waits most of the time, let’s spread its work over several processes, and assume that while most of the processes will be waiting, at least one can be free to deal with the local database. This way we will get maximum utilization of our resources.

So the multi-process workflow:

1. START TRANSACTION locally.
2. Fork children as necessary.
3. From child, make XML-RPC request, fetch data.
4. From child, acquire database access through semaphore.
5. From child, insert data into db as needed.
6. From child, release database access through semaphore.
7. From child, repeat 3-6 until Error or Finish.
8. From parent, monitor children until Error or Finish.
9. From parent, if Error ROLLBACK, if FINISH COMMIT.

Now, the workflow seems all and well in theory, but can it work in practice? Can we connect to the same transaction from several different PHP processes?

I was surprised to find out that the answer is positive. As long as all processes share the same connection resource, they all use the same connection. And in MySQL, the same connection means the same transaction, given that a transaction was started and not yet committed or rolled back (either explicitly or implictly).

The secret is to create the connection resource with the parent, and when forking children, they have a reference to the same connection. The caveat is that they must access the resource atomically, otherwise unexpected behavior occurs (usually the connection hangs, I am guessing that it is when one child tries to read() from the socket and the other to write() to it). So in order to streamline the access to the db connection, we use a semaphore. Each child can access the connection only when it’s available, and it’s blocking if not available.

In the end of the workflow, our parent process acts much like a Transaction Manager in an XA Transaction, and according to what the children report, decides whether to commit or rollback.

Here is a proof of concept code (not tested in this version, but similar code tested and succeeded):

The DBHandler Class

class DBHandler
{
	private $link;
	private $result;
	private $sem;
 
	const SEMKEY = '123456';
 
	public function __construct($host, $dbname, $user, $pass, $new_link = false, $client_flags = 0)
	{
		$this->link = mysql_connect($host, $user, $pass, $new_link, $client_flags);
		if (!$this->link)
			throw new Exception ('Could not connect to db. MySQL error was: '. mysql_error());
		$isDb = mysql_select_db($dbname,$this->link);
		if (!$isDb)
			throw new Exception ('Could not select db. MySQL error was: '. mysql_error());
	}
 
	private function enterSemaphore()
	{
		$this->sem = sem_get(self::SEMKEY,1);
		sem_acquire($this->sem);
	}
 
	private function exitSemaphore()
	{
		sem_release($this->sem);
	}
 
 
	public function query($sql)
	{
		$this->enterSemaphore();
 
		$this->result = mysql_unbuffered_query($sql, $this->link);
		if (!$this->result)
			throw new Exception ('Could not query: {' . $sql . '}. MySQL error was: '. mysql_error());
		if ($this->result === true)
		{
			// INSERT, UPDATE, etc..., no result set
			$ret = true;
		}
		else
		{
			// SELECT etc..., we have a result set
			$retArray = array();
			while ($row = mysql_fetch_assoc($this->result))
				$retArray[] = $row;
			mysql_free_result($this->result);
			$ret = $retArray;
		}
 
		$this->exitSemaphore();
 
		return $ret;
	}
 
	public function beginTransaction()
	{
		$this->query('SET AUTOCOMMIT = 0');
		$this->query('SET NAMES utf8');
		$this->query('START TRANSACTION');
	}
 
	public function rollback()
	{
		$this->query('ROLLBACK');
	}
 
	public function commit()
	{
		$this->query('COMMIT');
	}
}

The Forking Process

$pid = 'initial';
$maxProcs = $argv[1];
if (!$maxProcs)
{
	 $maxProcs = 3;
}
$runningProcs = array(); // will be $runningProcs[pid] = status;
define('PRIORITY_SUCCESS','-20');
define('PRIORITY_FAILURE','-19');
 
try 
{
	$dbh = new DBHandler(DBHOST,DBNAME,DBUSER,DBPASS);
 
	$dbh->beginTransaction();
 
		// fork all needed children
		$currentProcs = 0;
		while ( ($pid) && ($currentProcs < $maxProcs))
		{
			$pid = pcntl_fork();
			$currentProcs++;
			$runningProcs[$pid] = 0;
		}
 
		if ($pid==-1)
		{
			throw new Exception ("fork failed");
		}
		elseif ($pid)
		{
			// parent
			echo "+++ in parent +++\n";
			echo "+++ children are: " . implode(",",array_keys($runningProcs)) . "\n";
 
			// wait for children
			// NOTE -- here we do it with priority signaling
			// @TBD -- posix signaling or IPC signaling.
			while (in_array(0,$runningProcs))
			{
				if (in_array(PRIORITY_FAILURE,$runningProcs))
				{
					echo "+++ some child failed, finish waiting for children +++\n";
					break;
				}
				foreach ($runningProcs as $child_pid => $status)
				{
					$runningProcs[$child_pid] = pcntl_getpriority($child_pid);
					echo "+++ children status: $child_pid, $status +++\n";
				}
				echo "\n";
				sleep(1);
			}
 
			echo "+++ checking if should commit or rollback +++\n";
			if (in_array(PRIORITY_FAILURE,$runningProcs) || in_array(0,$runningProcs))
			{		
				echo "+++ some child had problem! rollback! +++\n";
				$dbh->rollback();
			}
			else
			{
				echo "+++ all my sons successful! committing! +++\n";
				$dbh->commit();
			}
 
			// signal all children to exit
			foreach ($runningProcs as $child_pid => $status)
			{
				echo "+++ killing child $child_pid +++\n";
				posix_kill($child_pid,SIGTERM);
			}
		}
		else
		{
			// child
			$mypid = getmypid();
			echo "--- in child $mypid ---\n";
			//sleep(1);
			echo "--- child $mypid current priority is " . pcntl_getpriority() . " ---\n";
 
			// NOTE -- following queries do not work, for example only
			$dbh->query("select ...");
 
			echo "--- child $mypid finished, setting priority to success and halting ---\n";
			pcntl_setpriority(PRIORITY_SUCCESS);
			while (true)
			{
				echo "--- child $mypid waiting to be killed ---\n";
				sleep(1);
			}
		}
 
} 
catch (Exception $e) 
{
	// output error
	print "Error!: " . $e->getMessage() . "\n";
 
	// if parent -- rollback, signal children to exit
	// if child  -- make priority failure to signal
	if ($pid)
	{
		// rollback
		$dbh->rollBack();
		foreach ($runningProcs as $child_pid => $status)
			posix_kill($child_pid,SIGTERM);
	}
	else
	{
		pcntl_setpriority(PRIORITY_FAILURE);
		$mypid = getmypid();
		while (true)
		{
			echo "--- child $mypid waiting to be killed ---\n";
			sleep(1);
		}
	}
 
}

Well, all of this sounds well, and also worked well on a development environment. But it should be taken out of the lab and tested on a production environment. Once I give it a shot, I will update with benchmarks.

It’s not a full blown phpMyAdmin, but it’s a simple way [...]]]> If you’re looking for a simple tool that uses Zend Framework’s robust database classes (such as Zend_Db and Zend_Db_Table), you can check out zdbform. It’s a short yet effective library that let’s you perform simple administration tasks on your database, with minimal coding.

It’s not a full blown phpMyAdmin, but it’s a simple way to view, edit and add your tables rows in a web interface. Also, don’t expect it to scale, because I am sure this library was written to serve some quick table administration needs, and that it is not ready to handle large datasets. But, it is very convenient if you have a small database to administer.

I implemented it with the Zend MVC components, and following is a brief overview.

In the front controller, or front plugin, or any class that your controller subclasses:

$this->db = Zend_Db::factory('Pdo_Mysql', array(
				'host'     => DB_HOST,
				'username' => DB_USER,
				'password' => DB_PASS,
				'dbname'   => DB_NAME
		));
Zend_Db_Table_Abstract::setDefaultAdapter($this->db);

Then set your controller and view scripts as necessary. Let’s say you have two tables to admin, “clients” and “history”. First make sure they are declared as subclasses of Zend_Db_Table:

require_once "Zend/Db/Table/Abstract.php";
 
class Clients extends Zend_Db_Table_Abstract
{
    protected $_name = 'clients';
}
 
class History extends Zend_Db_Table_Abstract
{
    protected $_name = 'history';
}

Your controller would look like:

require_once "Zend/Controller/Action.php";
 
class AdminController extends Zend_Controller_Action
{
 
	public function init()
	{
		require_once 'zdbform/zdbform.class.php';
		require_once 'zdbform/zdbform_widgets.class.php';
		require_once 'zdbform/zdbform_validations.php';
 
		parent::init();
 
		$this->view->headLink()->appendStylesheet('/zdbform/zdbform.css');
		$this->_helper->viewRenderer('index');
	}
 
	public function indexAction()
	{
	}
 
	public function clientsAction()
	{
		$this->view->dbform = new Zdbform('Clients');
		$this->view->dbform->setWidget('description', 'textarea');
		$this->view->dbform->processForms();
	}
 
	public function historyAction()
	{
		$this->view->dbform = new Zdbform('History');
		$this->view->dbform->processForms();
	}
 
}

And the single view script you need is admin/index.phtml:

<?php
echo $this->headLink();
 
include_once "Zend/Filter/Word/CamelCaseToDash.php";
include_once "Zend/Filter/Word/CamelCaseToUnderscore.php";
$cctd = new Zend_Filter_Word_CamelCaseToDash();
$cctu = new Zend_Filter_Word_CamelCaseToUnderscore();
$classes = get_declared_classes();
 
foreach ($classes as $class)
{
	if (is_subclass_of($class,'Zend_Db_Table_Abstract'))
	{
	?>
		<a href="/admin/<?= strtolower($cctd->filter($class)) ?>"><?= strtolower($cctu->filter($class)) ?></a>&nbsp;&nbsp;
	<?php
	}
}
 
if ($this->dbform)
{
	?>
	<h1>Table: <?= $this->dbform->tableName ?></h1>
	<?php
	$this->dbform->showForms();
	$this->dbform->showTable();
}
?>

There were also a couple things that needed changing in the zdbform class itself:

• Replace all PHP_SELF with REQUEST_URL. On the mvc case, PHP_SELF is empty or index.php, and we don’t want all the forms posted there, we want them to go back to /admin/clients or /admin/history
• After this line

  $this->pk = $tableInfo['primary'];

  I had to add this:

  if (is_array($this->pk))
  	$this->pk = $this->pk[1];

• zdbform->orderBy is treated as a single column, of you want multiple column sorting you have to hack a bit with getAllRows().

That’s it, point your browser to /admin and you’re good to go. In a very short time and with a little bit of code, you can get something similar to a stripped down version of phpMyAdmin, using the power of Zend Framework.

I usually live with the axiom that whatever you can find in the realm of Windows and proprietary software, you can easily find in the realm of Linux (any flavor) and open source. While this is indeed usually the case, when it comes to a Flex IDE for Ubuntu, there’s a real gap. Adobe has [...]]]>

I usually live with the axiom that whatever you can find in the realm of Windows and proprietary software, you can easily find in the realm of Linux (any flavor) and open source. While this is indeed usually the case, when it comes to a Flex IDE for Ubuntu, there’s a real gap. Adobe has their Flash IDE for willing and paying Windows users, and I am happy to say that I was one of these happy customers a while ago. But since then, Ubuntu has taken over my life, and when I set out to make a small Flex app a couple of days ago, I came across some hurdles. Not impossible to overcome, but not trivial as well.

The Options

There is no one complete solution for developing Flex on Linux. Many folks are looking for one, but there is still none to be found. There are many open source tools that cover vast areas of the ActionScript and SWF world, most of the listed on the wonderful osflash.org. Some of them are just right for you if you have a very specific task (like converting between formats, just compiling a bit of code, etc.), but none provide a complete IDE that lets you both drop in WYSIWYG elements and manually code some stuff, while easily maintaining complete control of what libraries are used.

The Choice

The choice then is to combine as few tools as possible. I have succeeded to get along with 2 tools: Flex Builder Linux Alpha, and Open Dialect.

Flex Builder Linux Alpha is an Adobe free product, which is a Flex build environment as a plugin for Eclipse. Don’t worry about the Alpha part, it seems like a very stable product, and besides eating up some memory, I had no problems with it. It is actually an exact replica of the Flex Builder for Windows, without the features of Design View, some wizards and the profiler.

Open Dialect is the most comprehensive attempt i’ve seen at creating a graphic WYSYWIG IDE for developing Flash. It has some basic characteristics of such an IDE, like a timeline with frames and key frames, and a graphic interface for creating shapes and editing their properties.

The Method

The development cycle is quite simple once you have these two tools running. Use Open Dialect for whatever graphic or animation you need, then grab the code from the “Document Script” tab, paste it to the Flex Builder in Eclipse, and start tweaking whatever is needed, add the MXML code etc. Open Dialect has great potential if it were to enable manual script editing, but currently it doesn’t.

Getting Things To Work

Requirements and Installation of Flex Builder Linux Alpha are covered on Adobe’s release notes. In short, you got to have Eclipse 3.3.x, Sun JRE 1.5.x and Firefox, and just follow the installation instructions there. Be sure to set Eclipse’s browser to Firefox, as mentioned in the release notes, and there are also a couple of guides to walk you through. Oh, and it’s very important to follow these instructions in order for your existing Eclipse plugins to survive the install. I installed the builder and lost my subclipse and collabnet merge plugin, and had to reinstall them.

Open Dialect uses the .NET framework, so you need Mono to run it. According to the installation instruction, you can either download pre-compiled binaries of Open Dialect, or download the source and compile, but then you need MonoDevelop as well. In my case, using apt-get install mono was enough, and Open Dialect ran like a charm.

Tweaking and Real Life Example

Let’s go through an example of how to make a rounded rectangle that gets its color through a FlashVar.

In Open Dialect:

1. Fire up Open Dialect.
2. Choose Rounded Rectangle from the Shapes list on the Items pane.
3. Draw a rounded rectangle on the canvas.
4. Set its X and Y to 0.
5. Go to the Document Script tab, select all the code and copy.

In Eclipse:

1. Fire up Eclipse, create a new Flex Project, name it “Test”, and go through the creation wizard (next, next, finish).
2. You are now editing Test.mxml which is a skeleton Flex app file. Paste everything you copied from Open Dialect into Test.mxml, instead of its current content.
3. Save.
4. Hey, we’ve got errors! That’s right, BListBox type is not defined. It’s because the script uses the “ActionScriptComponents” library that comes with Open Dialect, in which BListBox is declared, but we haven’t imported it yet, let’s do it.
5. Copy the directory /path/to/opendialect/ActionScriptComponents to /path/to/workspace/Test/src/
6. Run again.
7. Viola! Rounded rectangle showing up!

Now that we’ve covered the basics, let’s see how to pass FlashVars to our app. In order to do so, we need to understand what the Flex Builder environment does in build time — besides compiling the SWF, it also takes the file called index.template.html found in the html-template directory, and compiles it to a file called Test.html in the bin-debug directory, then fires this file in Firefox. So to pass FlashVars and process them in the script:

1. Open index.template.html
2. Scroll down to the javascript part under “hasRequestedVersion”, this is the part that runs the swf on our page (assuming we have javascript enabled and the correct version of the flash player).
3. Under

   "src", "${swf}",

   add

   "FlashVars", "color=0x000000",

4. Run once to see everything is working, and that we did not screw up the html template.
5. Add the following variable declaration where all other variable declarations are:

   private var rectColor:Number;

6. In init(), add the following line at the top of the function:

   rectColor = Application.application.parameters.color;

7. In the next line, where the shape properties are defined, change the last 2 colors (which in my case were 0×000000,0xFF0000) to rectColor,rectColor:

   RRect1A.Properties[0] = new ShapeProperties(0,0,120,172,"BRoundedRectangle",1,rectColor,rectColor);

8. Run again, see the rectangle is now black!

Conclusion

The field of Flex development on Linux is of course bound to change as time goes by, but for now it seems like it is still in its early unstable days. This was a brief demonstration of how to harness the IDE power of Open Dialect, and the development and build power of Adobe’s Flex Builder Linux Alpha, to create a working environment for developing Flex apps on Linux. It is by no means the simplest solution — my guess is that running Flex Builder on a Windows VMware could be easier, albeit costly — but this solution that I presented is of course free and conforms to the open source spirit. And, most importantly, can get you developing flex apps quickly and easily.

Have fun flexing!

A Couple of Sidenotes

• You don’t have to use the ActionScriptComponents library that comes with Open Dialect if you don’t intend to use tweening or frame events. You can use the regular mx.controls classes for just plain drawing.
• Flex applications tend to be big because the whole framework is included in the build. You can use RSL loading to cache the framework on the client side, or you can make a plain ActionScript Project instead of a Flex Project, still use mxml and just import the needed flash.* libraries. Both of these subjects are out of the scope of this post.

Internet Explorer expects a minimum sized [...]]]> Internet Explorer is a term you can’t stay indifferent to. Whenever you hear somebody say it, they either say it with a respectful awe, or they utter it with a developer’s pain. Today I was on the latter side of this indifference, when I fought a Wordpress bug.

Internet Explorer expects a minimum sized response content when receiving HTTP errors in the 400-500 area. If the content size received in the response is lower than this minimum, an arguably “Friendly HTTP Error Message” takes over, and displayed instead of the content.

This (rather obtrusive) behavior doesn’t go well with Wordpress wp_die() function, which is invoked on many occasions throughout the Wordpress code. This function sets the HTTP response status to 500, and then sends a very short html code containing any error message you’d like. These error messages are usually no more than a brief sentence, which results in a very small response content.

So take an IE7 browser, and try to post an empty comment on any Wordpress that is not upgraded to version 2.7 yet. See that friendly browser message? The same in Firefox produces the expected “Error: please type a comment.” message. Wordpress had already fixed this in 2.7 (you can copy their workaround) after some bug reports, but any installation prior to 2.7 which is still not upgraded (and many MU installations which always trail behind in the version releases) has this.

For any application other than Wordpress, be sure to go over this table of minimum content size for each HTTP status error code, and make sure your responses are long enough.

Apparently, sometimes being laconic is bad.

And then [...]]]> When it comes to client side, Javascript frameworks are one of the first things that accelerates your development of web applications. Before knowing about the existence of Javascript frameworks, I was coding raw Javascript, in strict functional programming, trying to solve the never ending war of cross-browser compatibility with every new line of code.

And then came along frameworks like Scriptaculous, JQuery, Dojo, and Mootools. These frameworks all tackle the everyday problems of Javascript (cross browser issues being the most important part), build and force an object oriented way of programming Javascript, and create easy to use classes for complicated tasks (like evaluating JSON responses from XmlHttpRequests). On top of that, they add cool and customizable graphics and effects for web UI.

History has its way of determining what framework you eventually use. They are all really similar, and when you’re just in the stage of picking your framework, it usually comes down to the one that has an example in its documentation that is most relevant to your current problem and seems that it solves it with the easiest code. Sometimes though, it’s just a matter of what framework is showcased better, and which has the coolest graphics.

In my case, it was Mootools, and I really can’t remember why. I started using it when it was in version 1.11, and I have been truely happy with it (except for a really nasty bug with https protocol and Internet Explorer). Mootools by version 1.11 was already a robust library, and didn’t require much more development. But of course that open source projects progress and develop, and have to catch up with new browsers and new technologies, and then Mootools released version 1.2.

It’s been already 6 months or so since it was released, and I still haven’t upgraded. And it’s not that I didn’t try — I have tried already 3 times I think — but the upgrade process from 1.11 to 1.2 is probably the hardest upgrade I’ve ever encountered. Version 1.2 is not backward compatible with version 1.11. There are compatibility packages, and several attempts of the community to build on top of these packages, but they never seem to cover all the compatibility needed. There are always a few lines of code you wrote using 1.11, that even with cmpatiblity packages, 1.2 will just break on.

Now don’t get me wrong, I am addicted to Mootools. They say that its core is stable, and they’re right. I just hope that I will never be forced to try again to upgrade to 1.2, or that there will be an easy drop-in way to get code written for 1.11 run on 1.2.