On Tue, 22 Nov 2011 14:49:37 +0100 "Rolf E. Sonneveld" <[email protected]> wrote:
> Is this sufficient 'evidence' that MD + stream_by_recipient()+ > Postfix (non_smptd_milters) provides a solution to the problem of > treating a multi-recipient message differently based upon different > 'per-user' preferences/settings? Any things I forgot? Any caveats? It looks like it's working. However, there's one caveat: With real Sendmail, MIMEDefang redelivers the streamed messages using deferred mode. That means they just get queued up. A short time later, the queue is run and the remailed messages appear. This means that if a message has 100 recipients, they get queued up and then redelivered in a nicely serialized way with limited parallelism. If Postfix actually redelivers the messages immediately, an N-recipient message might try to tie up N scanning processes all at about the same time. I'm not sure if this will be a problem in practice, but it's something to watch for. You don't want to allow an attacker to DoS your machine by sending messages to large numbers of recipients and relying on amplification. Regards, David. _______________________________________________ NOTE: If there is a disclaimer or other legal boilerplate in the above message, it is NULL AND VOID. You may ignore it. Visit http://www.mimedefang.org and http://www.roaringpenguin.com MIMEDefang mailing list [email protected] http://lists.roaringpenguin.com/mailman/listinfo/mimedefang
