At 9:37 PM +0100 3/29/01, G.W. Haywood wrote:
>Hi there,
>
>On Thu, 29 Mar 2001, Robert Landrum wrote:
>
>> (once a week maybe), I get a string of error reports
>[snip]
>> Can't locate auto/CapWiz/Mail/App/handler.al in @INC
>[snip]
>
>This puzzles me...
>
>> at /dev/null line 0
>
>Are you getting segfaults?  Running short of memory?
>Any other clues?

I haven't been seeing any core dumps.  My apache processes are rather 
large to begin with (30MB).  I wish I had more to go on, but all I 
get are the occasional error report to my webmaster account.  It 
seems very odd.


> > The CapWiz::Mail::App has a handler
>
>What and where is it?


/usr/local/lib/perl5/site_perl/5.6.0/CapWiz/Mail/App.pm

We (my company) send email to congress (for napster and others). 
CapWiz::Mail::App is the Application that composes and sends all the 
email.

Here is what the handler looks like.  It is the first sub in the App.pm file.

sub handler {
         my $r = shift;
         my $app = CapWiz::Mail::App->new();
         return $app->run($r);
}

We don't do anything weird... I also have a 'use CapWiz::Mail::App;' 
in my startup.pl file.  As far as I can tell, there is no way that 
this could happen on a normally loaded server.  I could see it 
happening on an overloaded server that has too many open files or 
runs out of memory, but our server has neither of these problems.

>
>> Has anyone else had these problems in the past?
>
>Not I.
>
>> Apache 1.3.12, Perl 5.6.0, mod_perl 1.24_01
>
>Might be worth trying a later Perl in the hope of getting a more
>informative/different error message but that's a shot in the dark.

Is there any way that it's apache?  I have yet to reproduce it on our 
development server which is Apache 1.3.14, perl 5.6.0, mod_perl 
1.24_01.

I may end up just upgrading the whole system to 1.3.19, 5.6.1, 1.25 
when 5.6.1 is stable...

Thanks,

Rob

--
As soon as you make something foolproof, someone will create a better fool.

Reply via email to