Batch process (+8000 pdf) help!

[Francisco Alvarez Padilla]

  Hi i want process more 8000 pdf in one process.

The report reaches the limit and the process is broken to reach the limit of memory allocated to that process, does not exceed 2 GB Transaction.

Now try both and load_html_file load_html and both thunders. (beta2 version)

dumpdf has some method to free the occupied memory to generate the pdf?

Someone has gone something like this?

plz help

gretings

[Fabien Ménager]

Hello, dompdf does have such a method, but it's called internally each time an object is not needed anymore. This method is not used in PHP 5.3 because this version has a good garbage collector. Which version of PHP do you use ?

Do you instanciate a new DOMPDF object each time ? Are there any PDF generated ? On of the document may stop the process because it has the issue evoqued here :  http://code.google.com/p/dompdf/issues/detail?id=91 

[Francisco Alvarez Padilla]

HI thaks for the answers

For each interaction i use that

        $dompdf = new DOMPDF();

$dompdf->load_html($html);  // $dompdf->load_html_file($urlFile);

$dompdf->render();

$pdfdata=$dompdf->output();

file_put_contents("../pdf/".$fileName.".pdf", $pdfdata);

unset($html); // unset($urlFile);

unset($dompdf);

The version of php is: PHP Version 5.3.5

The server for testing ant take care that the process is done with xampp

Also start the memory 

ini_set('memory_limit', '512M');

I the log errors have this

PHP Fatal error:  Allowed memory size of 536870912 bytes exhausted (tried to allocate 393216 bytes

Only can create 350 pdf, of 8000

any one can help me?

Thx!! Gretings.

[Fabien Ménager]

Could you write some sort of memory logging after each step :

• before new dompdf
  
• after load_html
• after render
• after the unsets

and write it in a log file with the current document name you are processing.

With this you would see if it is a regular memory usage growth or if a document takes all the memory because it has the problem I mentionned.

[Francisco Alvarez Padilla]

Hi

I get this results... (some)

the value for render, dont print nothing, i dont understand why.

Before: 1989976

In load_html_file: 2340144

unset: 9672328

Before: 9672904

In load_html_file: 9676976

unset: 10168088

Before: 10166272

In load_html_file: 10170376

unset: 10542760

Before: 10542728

In load_html_file: 10546768

unset: 10913328

Before: 10913504

In load_html_file: 10917672

unset: 11267656

Before: 11267640

In load_html_file: 11271680

unset: 11629376

Before: 11629632

In load_html_file: 11633672

unset: 12000672

The log File dont generate any error.

Gretings

[BrianS]

PHP does do some amount of garbage handling, but some of this just indicates that previously-used memory is no longer in use. Ideally PHP would release or reuse the memory. But that's not always the case. What we've found in the past is that the most efficient method of running DOMPDF in batch is to run each rendering in a separate PHP process via exec(). Check out the following discussion:

https://groups.google.com/d/topic/dompdf/wlOlKxB2ajM/discussion

[Francisco Alvarez Padilla]

Hi Brian Thank so much for the anwser.

In the info for usage,exist an a example for this?

Thanks!

[BrianS]

This isn't something we've really documented (or even thoroughly researched) just yet. So unfortunately there isn't much info in the dompdf documentation. If you need any advice on how to proceed let us know.

[Francisco Alvarez Padilla]

Ups, ok, ahhh what is the better way for this?

Create a .bat file with my process, and execute with exec. This .bat process a php with dompdf

OR

Make testing executing a dompdf command line (i see a little documentation) if sucess, make the exec process. 

Thanks Brian!

Gretins

[BrianS]

You can keep your current file, but instead of something like this:

...
foreach ($htmldoc_collection as $htmldoc) {
  $dompdf = new DOMPDF();
  $dompdf->load_html($htmldoc);

  $dompdf->render();
  $pdfdata=$dompdf->output();
  file_put_contents("../pdf/".$fileName.".pdf", $pdfdata);

  unset($dompdf);
}
...

you would do something like this:

...
foreach ($htmldoc_collection as $htmldoc) {
  file_put_contents("doc.html", $htmldoc);
  exec("php pdfgen.php doc.html $filename");
  unlink("doc.html");
}
...

pdfgen.php would look something like this:

$htmldoc = $argv[1];
$filename = $argv[2];
$dompdf = new DOMPDF();
$dompdf->load_html_file($htmldoc);
$dompdf->render();
file_put_contents($fileName, $dompdf->output());

Modify the above as necessary to get things working how you want, it may take some trial-and-error.