Re: SMTP traffic control
2011-10-28 19:55:08
Carl S. Gutekunst wrote:
Hector Santos wrote:
Unfortunately, the dominant conversation seems to be around an SMTP
response that only applies to greylisting. I'd never use that or
support it, it's such a teeny-tiny corner case, especially in the
B-to-B world where I live.
Does this suggests that your B2B market is oblivious to the outside
world remote server behaviors employing 4yz temporary rejections?
They're highly conscious of 4xx temporary rejections associated with
resource exhaustion and down systems. But in 20 years of supporting
B-to-B E-mail, I can only remember one issue with someone else's
greylisting -- and that was because it was broken.
Ok, and we should|must all respect that. Nevertheless, its not the
case for all though, nor should it be reason for not considering this.
It just means it doesn't apply to your market area.
My point is if like Atkins proposal and might implement it, once you
do, you will be behaving like a Greylisted server.
Most better SMTP servers already deal with DoS related issues and
employ a concept for loading limiting/restrictions. What has been
seen with some real servers 421 responses today, and now also being
suggested is adding "time delay" hints to this load limiting 421
rejections because it is also behaving like a Greylisting server.
Again, for lack of a better term, Greylisting, is really two parts:
- Using 4yz for unknown, non-whitelisted, unrecorded senders
immediately,
- Blocking them for a specific, implementation based time length.
If we remove the blocking time from the equation, i.e. it was never
part of the logic, then I don't believe we would be having this
discussion. All 2nd tries will work and delays will be limited to the
MTA's 2nd attempt time.
But thats not the case. The issue is that is there are blocking times
and in comes in varying degrees and implementing Atkins proposals
pushing you into something you indicated you don't wish to support.
Look, I think both proposals are ok. We do need something that
documents Greylisting because its more than just issuing a "time
delay", it about monitoring and tracking and using blocking times.
I could change the SMTPGREY proposal to use "wait=" which would
reference a more "Generic" proposal on Server/client waiting times.
I would rather that we work together to seek a common solution to all
the issues and needs, rather than begin starting and playing this
partisanship games once again.
--
Sincerely
Hector Santos
http://www.santronics.com
<Prev in Thread] |
Current Thread |
[Next in Thread>
|
- Re: SMTP traffic control, (continued)
- Re: SMTP traffic control, Tim Kehres
- Re: SMTP traffic control, Hector Santos
- Re: SMTP traffic control, Carl S. Gutekunst
- Re: SMTP traffic control,
Hector Santos <=
- Re: SMTP traffic control, Steve Atkins
- Re: SMTP traffic control, Carl S. Gutekunst
- Server Enforcement of Time Blocks (wait=), Hector Santos
- Re: Server Enforcement of Time Blocks (wait=), Steve Atkins
- Re: Server Enforcement of Time Blocks (wait=), Hector Santos
- Re: Server Enforcement of Time Blocks (wait=), Peter J. Holzer
- Re: Server Enforcement of Time Blocks (wait=), Hector Santos
- Re: Server Enforcement of Time Blocks (wait=), Peter J. Holzer
- Re: Server Enforcement of Time Blocks (wait=), Carl S. Gutekunst
- Re: Server Enforcement of Time Blocks (wait=), Hector Santos
|
|
|