On Wed, Apr 15, 2009 at 10:29 AM, Iain <[email protected]> wrote:
> I discovered by accident (by impatiently trying to read an email that I knew
> was being delivered by getmail into Sup) that I can reliably replicate the
> problem, or at least a similar problem. It seems to occur when Sup polls the
> Maildir at the same time as getmail is retrieving the message.

I thought this might be the problem with my lost messages.  So I
changed my setup so that I don't have fetchmail running as a daemon
when sup is running.  Instead, I have sup running fetchmail whenever
it polls.  But I'm still getting lost messages, maybe 5 per day (out
of hundreds of message).  So there may be two problems here.

I've started writing some debugging code to see if I can track down
these lost messages.  Attached is a little program that parses a
procmail log file, and checks to see if each message mentioned in the
log also appears in the ferret index.  It's got some ugly hacks for my
particular setup (e.g., the checks for the "vd" and "rd" directories,
which I need to ignore in my case, and the hardcoded Maildir
directory path).  But maybe it'll prove useful to
others who are using procmail.

I invoke this script from the sup source directory using:

  ruby -I lib procmail.rb /home/marka/Maildir/procmail.log

I have this in my ~/.procmailrc to enable logging:

  LOGFILE=/home/marka/Maildir/procmail.log

Attachment: procmail.rb
Description: Binary data

_______________________________________________
sup-talk mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/sup-talk

Reply via email to