Re: Extended 5xx code (was Re: Drop UCE instead of forwarding off-site?)
- From: "D. Stussy" <spam@xxxxxxxxxxxxxxxx>
- Date: Sat, 16 May 2009 18:12:41 -0700
"Xavier Roche" <xroche@xxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:gumh7j$b7r$1@xxxxxxxxxxxxxxxxxxx
David F. Skoll a écrit :friendly
No, it doesn't validate the suggestion. If the back-end server
(a) thinks something is spam, and (b) realizes it's coming from a
server, it should silently discard it rather than rejecting it.
Humm, if the mail was actually not a spam, you are:
(1) violating the SMTP protocol
(2) destroying a private mail, with potential consequences (legal, term
of service issues with your clients, and so on)
Isn't filtering with a back-end server a design error, since there is no
reliable solution to (1) avoid backscatters and (2) respect the SMTP
protocol (and not be liable for silent mail drops) at the same time ?
ALL MTAs should filter. Due to race conditions, it is possible that a
message accepted (therefore, not spammy) by a front-end MTA is known to be
spammy by the time it is forwarded to a back-end MTA, especially when
external databases (be they DNSBLs, checksum based, etc.) are used to
classify spamminess.
The proposal for "ExtendedErrorDrop" is so that a back-end server can tell
a front-end that a drop is preferred to generating an NDR for CERTAIN types
of rejections (based on the DSN returned, and ONLY those DSNs indicated by
the proposed option).
[I have never been convinced that having "N" frontend servers accepting
mails + "M" backend servers filtering mails is better than having "N+M"
frontend servers filtering mails directly -- two passes are both less
reliable and require more resources at the end]
Nowhere did I suggest that the front-end not filter. In fact, see above.
.
- Follow-Ups:
- Re: Extended 5xx code (was Re: Drop UCE instead of forwarding off-site?)
- From: Xavier Roche
- Re: Extended 5xx code (was Re: Drop UCE instead of forwarding off-site?)
- References:
- Drop UCE instead of forwarding off-site?
- From: Hauke Fath
- Re: Drop UCE instead of forwarding off-site?
- From: D. Stussy
- Extended 5xx code (was Re: Drop UCE instead of forwarding off-site?)
- From: David F. Skoll
- Re: Extended 5xx code (was Re: Drop UCE instead of forwarding off-site?)
- From: Xavier Roche
- Drop UCE instead of forwarding off-site?
- Prev by Date: Re: Extended 5xx code (was Re: Drop UCE instead of forwarding off-site?)
- Next by Date: Re: Request for feature - Option "ExtendedErrorDrop" => do not generate NDR.
- Previous by thread: Re: Extended 5xx code (was Re: Drop UCE instead of forwarding ?off-site?)
- Next by thread: Re: Extended 5xx code (was Re: Drop UCE instead of forwarding off-site?)
- Index(es):
Relevant Pages
|