Author: Evgeniy Berdnikov Date: To: exim-users Subject: Re: [exim] Special log after mail has been delivered
On Tue, May 07, 2013 at 01:52:34PM +0200, Heiko Schlittermann wrote: > Evgeniy Berdnikov <bd4@???> (Di 07 Mai 2013 09:37:52 CEST):
> > On Mon, May 06, 2013 at 10:02:35PM +0200, Heiko Schlittermann wrote: ... > Rejecting a message is not bouncing it. Bouncing a message is generating
> a new message.
OK, terminology fixed. :)
> > > I'd say, the built-in logging facility of Exim is much more mature,
> > > since it stops message reception in face of logging problems. IMHO a
> > > very crucial element of a safe and secure mail service.
> >
> > Agree. However, writing to named pipe is so safe as writing to a file:
> > if nobody reads pipe, then writer is blocked.
>
> Agree, BUT??? I'd say in case of Exim it is a difference, because Exim
> checks the available space on the log file system. The available space
> is only relevant, if we're really writing into files. If Exim writes
> into a named pipe, we have one more level of complexity???
>
> (see check_log_{space,inodes})
Writing to database inevitably leads to one more level of complexity.
Well, you can't use check_log_{space,inodes} if you write to database
on the fly, but if write to pipe blocks then exim also hangs and no mails
would be accepted. Surely, it's not so good as built-in space check, but
result is effectively the same. In any case, on the edge exim->database
error conditions must be held carefully: db client should stop reading
data if it can't update+commit, and it should block feeder if possible.
--
Eugene Berdnikov