Ok, two things.. whether or not QS detected the right error, it's going
to exit with 111. I see error 111 is a soft failure, retry later. 
(According the 'The Big Picture')
Would email clients not heed that error?  I did a quick Google search,
but I'm getting just about everything with an error 111 :)

It almost looks like 111 is the best code that's available, and just
'test the client' ;)

Rick



On Tue, 2003-02-25 at 15:12, Rick Romero wrote:
> Ok, here's the deal.. I run FreeBSD, originally 4.3-stable.
> 
> I did a complete ports update, but I have never done a 'make world'.  I
> want to test the latest Horde/IMP, which required PHP 4.2.1.  So I
> started updating the required ports to do so. Apache, gettext, etc.  I
> don't know what package reformime and Apache/PHP require in common, but
> the end result was a broken reformime.
> 
> I run Qmail-Scanner-1.13, and I see in the code, that my situation
> SHOULD be accounted for:
> -----------------
> &debug("d_m: starting $mimeunpacker_binary
> <$scandir/$wmaildir/new/$file_id [",&deltatime,"]");
>   open(MIME,"$mimeunpacker_binary 
> <$scandir/$wmaildir/new/$file_id|")||&tempfail("cannot call mimeunpacker
> - $!");
>   while (<MIME>) {
>     next if (/exists\./);
>     &tempfail("d_m: output spotted from reformime - that shouldn't
> happen!");
>   }
>   close(MIME)||&tempfail("cannot close mimeunpacker - $!");
>   my $unpacker='';
> -------------------
> 
> Unfortunately, my maillog has a bunch of:
> -------------------
> Feb 25 00:10:50 www X-Qmail-Scanner-1.13:   
> [www.myhost.net104615345042369279] cannot close mimeunpacker -
> -------------------
> 
> Now, when I ran reformime from the command line (when it was broken), I
> got an error.  Sorry, I don't remember what it was, and it didn't make
> it into any logs.  Not being a PERL guy I think:    
> next if (/exists\./);
> checks for a period in the output.  I don't think there was a period in
> the error.  IIRC, it was some kind of lib issue.  But there were plenty
> of capitalized letters ;)
> 
> Can the length maybe be checked instead of a period?
> 
> I really apologize for not being able to provide more than a report..
> but I bet I'd probably spend hours trying to figure out a good QUICK
> 'SOMETHING' that someone else out there already knows :)
> 
> Now, I'm REALLY not complaining.. I realize *I* broke it in the first
> place.  But, at the same time, the mail shouldn't have been lost.  Have
> I screwed something else up, so that the 'exit 111' in QS isn't
> triggering something right? 
> Even more strange, only 'outgoing' mail was affected.  Anything coming
> in (such as mailing lists to users), was deferred on the remote server.
> Only mail submitted by an email client was lost.  (I sent tests to
> myself, and NEVER received them, but I eventually got a post to a
> mailing list I made during the 'downtime' via another account.)
> 
> I run QS via the qmail-queue patch, so it should be called on ALL
> messages.  
> 
> Other than nice log of any messages the server recieved, vs a client -
> which I don't have :( is there anything else that might help?
> 
> Rick
> 
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Qmail-scanner-general mailing list
> [EMAIL PROTECTED]
> https://lists.sourceforge.net/lists/listinfo/qmail-scanner-general
-- 
--
Rick Romero
IT Manager      
Valeo, Inc.     ph: 262.695.4841
Sussex, WI.     fax: 262.695.4850
[EMAIL PROTECTED]



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Qmail-scanner-general mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/qmail-scanner-general

Reply via email to