I'm probably not the most qualified person to talk about this ... but...
Duplicate elimnation is not a runtime or build option. It is only
disabled in the event of some strange failure in the initialisation of
the database (no permissions on the /var/imap/deliverdb directory or
similar).
If you look at (2.0.16) imap/lmtpd.c line 281...
/* initialize duplicate delivery database */
dupelim = 1;
if (duplicate_init(0) != 0) {
syslog(LOG_ERR,
"deliver: unable to init duplicate delivery database\n");
dupelim = 0;
}
"dupelim" is a global variable. It is tested whenever the duplicate
deliver mechanism is to be invoked. You could try commenting out the
call to "duplicate_init()" and make sure dupelim is set to 0.
However, it is probably worth checking that dupelim is used
consistently...
If it does work, then the thing to do would be to add a config to
imapd.conf to control wether or not you want duplicate elimination.
Hope that helps.
Ian.
PS. 1,500,000 mailboxes does sound interesting!
On Wed, 2001-11-14 at 08:19, Birger Toedtmann wrote:
>
> Hi,
>
> within the Cyrus IMAP package (2.0.16) there is much talk about sieve only
> working with duplicate delivery suppression and disabling that would also
> disable sieve.
>
> No more information is supplied, only the configure switch --disable-sieve.
>
> My point is: I need *fast* delivery (you may find out why in a different
> post "Patch: config file for master as argument" on this list). I don't
> need sieve, and I don't need the duplicate delivery suppression code.
>
> What I did not find is: how can I disable that code? Is there a "hidden"
> switch in configure? (read the source to no avail)
>
> When I install Cyrus IMAP with most features turned off, I'll still get
> the "duplicate-X".db files and Cyrus working on them, which wastes my
> ressources (it boils down to that - I know that duplicate delivery supp-
> ression and sieve are good ideas in general, but I cannot use them in the
> situation at hand).
>
> Your ideas?
>
>
> Regards,
>
> Birger
>
