Handling non-fatal "bounces"
So we converted the firewalls mailing list to mailman, and we're now being inundated by bounce messages that mailman isn't handling. I'm going to write bounce modules for the actual bounces, but I can't figure out how to handle non-fatal errors (message could not be delivered, will retry soon, etc.).
What I want to do is ignore the bounce (don't flag the list member), but _not_ forward the message to the list admins. The current API seems incapable of doing this. Am I missing something?
-- Carson Gaspar - carson@taltos.org Queen trapped in a butch body
"CG" == Carson Gaspar <carson@taltos.org> writes:
CG> So we converted the firewalls mailing list to mailman, and
CG> we're now being inundated by bounce messages that mailman
CG> isn't handling. I'm going to write bounce modules for the
CG> actual bounces, but I can't figure out how to handle non-fatal
CG> errors (message could not be delivered, will retry soon,
CG> etc.).
CG> What I want to do is ignore the bounce (don't flag the list
CG> member), but _not_ forward the message to the list admins. The
CG> current API seems incapable of doing this. Am I missing
CG> something?First off, I consider it very bad manners for a side to send non-fatal warnings (i.e. retries) for messages coming from a list. IOW, if the message has "Precedence: bulk" a site should /not/ send a try notification. Most MTAs I believe either suppress these by default, or can be configured to do so. I seem to remember a discussion concerning an Exim configuration variable to control this a while back (perhaps Nigel can elaborate).
In general, I haven't added bounce handlers to detect these kinds of messages, and you're correct that the interface in BouncerAPI.py doesn't really support an "ignore this, they're bogus" return value. Usually when I'm bombarded by these, I send a, er, nicely worded message to the postmaster and the warnings usually stop.
It'd be fairly easy to extend BouncerAPI.py to handle ignorable bounces, so I'll try to add something for 2.1.
One question: should Mailman support doing an (optional) auto-kvetch back to the site's postmaster?
-Barry
On Mon, Jun 18, 2001 at 05:21:19PM -0400, Barry A. Warsaw wrote:
One question: should Mailman support doing an (optional) auto-kvetch back to the site's postmaster?
Two words: "Sorcerer's Apprentice".
Cheers, -- jr 'IE: "no" :-)' a
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
OS X: Because making Unix user-friendly was easier than debugging Windows
On Monday, June 18, 2001, at 02:42 PM, Jay R. Ashworth wrote:
On Mon, Jun 18, 2001 at 05:21:19PM -0400, Barry A. Warsaw wrote:
One question: should Mailman support doing an (optional) auto-kvetch back to the site's postmaster?
Two words: "Sorcerer's Apprentice".
treat them as hard bounces and if someone gets unsubscribed, encourage them to whine at their admin.
sending stuff to the admin directly from the outside will only piss off the admin. it won't solve anything, and you're kvetching on a point that is arguable depending on who's side of the argument you're on (if you want to kvetch at things, kvetch at places that send stupid vacation messages to the entire universe, endlessly, or places that send "confirming we read your message" messages to mail lists. Those are a LOT less ambiguously screwed up than the not-delivered-yet pain-in-the-necks...
-- Chuq Von Rospach, Internet Gnome <http://www.chuqui.com> [<chuqui@plaidworks.com> = <me@chuqui.com> = <chuq@apple.com>] Yes, yes, I've finally finished my home page. Lucky you.
I'm really easy to get along with once you people learn to worship me.
On 18 Jun 2001 17:21:19 -0400, Barry A. Warsaw wrote:
First off, I consider it very bad manners for a side to send non-fatal warnings (i.e. retries) for messages coming from a list. IOW, if the message has "Precedence: bulk" a site should /not/ send a try notification. Most MTAs I believe either suppress these by default, or can be configured to do so. I seem to remember a discussion concerning an Exim configuration variable to control this a while back (perhaps Nigel can elaborate).
Sure.
Exim can send delay warning messages out at configurable intervals - see http://www.exim.org/exim-html-3.20/doc/html/spec_11.html#SEC201
A condition variable also controls what types of messages do get delay warnings sent out. By default this does not send warnings out in response to messages with Precedence: set to bulk, list or junk.
Older versions of exim (I guess pre-3.x - which are prehistoric now) did not have the defaults appropriately set. If you are using a version of exim that old then you should upgrade, or if this is impossible then set delay_warning_condition appropriately - documentation for this (from the 3.2x documentation - but this hasn't changed since its original introduction) is at http://www.exim.org/exim-html-3.20/doc/html/spec_11.html#SEC202
It'd be fairly easy to extend BouncerAPI.py to handle ignorable bounces, so I'll try to add something for 2.1.
That would be good. Is there a way of hacking it now? [return an impossible address for example]
One question: should Mailman support doing an (optional) auto-kvetch back to the site's postmaster?
My evil twin says yes to this... however since those people that have their systems so badly configured that they send bounce messages for list mail also don't respond to postmaster mail - one site that is causing me particular problems in that respect currently I now blacklist completely. Autoresponses in error conditions are a good way of building serious mail loops.
Nigel.-- [ Nigel Metheringham Nigel.Metheringham@InTechnology.co.uk ] [ Phone: +44 1423 850000 Fax +44 1423 858866 ] [ - Comments in this message are my own and not ITO opinion/policy - ] [ ----- Security is not an add-on -- security is a way of life ----- ]
participants (5)
-
barry＠digicool.com -
Carson Gaspar -
Chuq Von Rospach -
Jay R. Ashworth -
Nigel Metheringham