I hope this doesn't sound like a vpopmail newbe question, but after some
tests I conducted (send 5000 messages to a /etc/passwd user takes 30 minutes
on my machine, the same 5000 messages to a virtual user through vpopmail
(working with cdb) takes 50 minutes), I am wandering, why does vpopmail uses
an executable file for final delivery in qmail-default of the domain,
instead of just putting a .qmail-username with the user's maildir to deliver
to in domain home directory ? I believe that this could decrease the
processing overhead.

I understand that this could hert some functionality, like per-domain quotas
(if someone need a per-user quota, he can put a quota-limit script in
.qmail-username, which would also allow for per-user quotas, which is a
needed functionality that seems to be lacking in vpopmail), and also the
root domain directory will be cluttered with .qmail files, but this is a
small price to pay for such a noticable performance boost.


Reply via email to