On Tue, 7 Oct 2003, Avleen Vig wrote:
> On Tue, Oct 07, 2003 at 05:54:35AM -0500, Jerry Jorgenson wrote:
> > Do I see an EximII.the.sequel here, with a separate code branch?
1. I do *not* want to go through another Exim3/Exim4 transition if I can
help it. Actually, since this is very internal, I doubt that such a big
change would be necessary. Anything this is done in this are could
almost certainly be done in a compatible manner.
2. This is a big item; I'm not going to be implementing anything big
very soon (away rather a lot and need to get documentation updated for
the next release). I also intend to initiate a discussion on the
long-term sometime next year (as I have posted a couple of times).
> 2) structure the db better.
> the current current db uses a simple key/value pair, where the
> key is the MX hostname and the value is a long string of message
> ID's waiting for delivery. part of the blocking is here; when I
> have multiple queue runners trying to write to the same key's
> value.. :-)
Yes, this is where the assumption of "small queue, not many delayed"
really bites.
> 3) it can even be beneficial to *not* use a hints db. maybe have an
> which defaults to having the db enabled?
I presume you mean not to use an smtp-wait hints db. I suspect not that
using a normal retry hints db would not be such a good idea.
An option not to use an smtp-wait db would be utterly trivial to
implement, so it would be easy to test. All that is needed is for a
couple of lines of code not to be obeyed.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
Get the Exim 4 book: http://www.uit.co.uk/exim-book
