On Wed, 2007-10-24 at 21:07 +0000, Werner Schneider wrote:
> Hi, I got a strange problem: Using php 4.4.x, I capture the whole output for
> a webpage into the output-buffer by using ob_start and ob_get_clean, because
> I got to make some replacements in the html-code before sending the page to
> the browser.
>
> This worked fine with a small page, but now I got a page for which the
> html-code is about 280 KB (not too big I think). But I get an server-error
> 500 for this script on my linux-based webhoster.
>
> I tried to run it on an local WAMP-installation - it worked without error.
> I temporarly deleted some of the output - it worked without error.
> I turned of the output-buffering - it worked without error.
>
> I printed memory_get_usage, but it never exceeded 2 MB (I load about 200
> records from a database and create objects from it and then print the data of
> these 200 records).
> On phpinfo(), memory_limit of my webhoster is 40M - so 2MB SHOULD be no
> problem.
> I increased the memory-limit to 64M - it doesn't help.
>
> Any idea what I could do next? I got no access to the apache error-log, but
> the error is definitely connected to the output buffer and how much I try to
> store in it.
>
> Any help is welcome.
Use some random URL parameter and detect it in your script. When
detected enabled display errors:
<?php
if( isset( $_GET['knjdcrksjhfcsjkhfndkf'] ) )
{
ini_set( 'display_errors', 1 );
}
?>
Then see if you get any errors. Although, if you're seg faulting, you
still won't see an error since the program just dies.
Cheers,
Rob.
--
...........................................................
SwarmBuy.com - http://www.swarmbuy.com
Leveraging the buying power of the masses!
...........................................................
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php