procmail
[Top] [All Lists]

Re: NULLs -- LOST EMAIL

1997-02-18 21:12:38
Responding To:  Philip Guenther <guenther(_at_)gac(_dot_)edu>
Original Date:  Tue, 18 Feb 1997 21:52:20 -0600
Message-ID:     <199702190352(_dot_)VAA19778(_at_)solen(_dot_)gac(_dot_)edu>

Was procmail
compiled on your system or did you try to get the binary from
someone else?

I compiled it on the system that it is currently using


Did you give enough directories to the locking tests?

All the mail goes on the same system, on one disk.  How many  
directories should I give it?  It would seem that this setup would  
not require more than one directory for locking tests, and if my  
setup is not the perfect setup for one directory, then what system  
is procmail's default (just using /tmp) designed for?


Do all the recipes in your .procmailrc that write to your mailspool
the the _correct_ locallockfile?

This I do not understand.  I have only two recipes which actually  
dump to $DEFAULT (the rest just fall off the end of the procmailrc),  
and both are used this way:

This is about halfway through my .procmailrc:

:0:
* ^Subject: (NeXT_FTP_Report|\
             NEXTSTEP/OpenStep Resources on the Net)
$DEFAULT

and this is at the very end

:0
{       
        :0cwhi: $PROCDIR/sndplay.lock
        * ! ? test -f $PROCDIR/hushmail
        |ismail

        :0cwhi: $PROCDIR/mesgconsole.lock
        |(/bin/echo "   From: $FROM2"  && \
          /bin/echo "Subject: $SUBJECT") > /dev/console


        :0:
        $DEFAULT
}       


Is this correct?  Here is my procmail -v info:

procmail v3.11pre4 1995/10/29 written and created by Stephen R. van  
den Berg  <srb(_at_)cuci(_dot_)nl>

Submit questions/answers to the procmail-related mailinglist by  
sending to:
        <procmail(_at_)informatik(_dot_)rwth-aachen(_dot_)de>

And of course, subscription and information requests for this list to:
        <procmail-request(_at_)informatik(_dot_)rwth-aachen(_dot_)de>

Locking strategies:     dotlocking, flock()
Default rcfile:         $HOME/.procmailrc
System mailbox:         /usr/spool/mail/$LOGNAME



With your mailreader, are you *sure* it has a locking strategy
in common with procmail?

My main reader only accesses the spool when I tell it to (manually)  
and none of this has ever happened when I was telling it to  
retrieve mail from the mailspool.  I can see when my POP binary  
starts to run, which passes the mail to procmail, and I can see when  
it exits (ie procmail is far done by the time I tell my mailreader  
to go for my mailspool).  Therefore I do not think that this has  
anything to do with my mail reader.

My POP program takes all the messages from my ISP and sends them to:

        |/usr/local/bin/formail -e -s procmail

Which seems to handle one message at a time (ie the log shows only  
one message being processed at a time as far as I can tell)



If I recompile procmail, what should I do differently?  I don't  
have any NFS mounted drives, so that shouldn't be a problem.  Even  
with one drive should I use multiple directories?  Is there a way to  
force it to act more conservatively with regard to locking?

You may have nailed the problem.  This is very sporadic, so I can  
only guess that the problem is intermittent, but for the life of me  
I can't find any common thread that might be the problem.

Thanks for your time

TjL


-- 
Tj Luoma (luomat(_at_)peak(_dot_)org) 
If you have a web page about NeXTStep|OpenStep, email me the URL!

EMAIL ADDRESS: Please use the PEAK address and not the NERC one

<Prev in Thread] Current Thread [Next in Thread>