Charlie Brady wrote:
Suppose you have two addresses that forward to each other, either both
on the local server or one remote and one local.
qpsmtpd doesn't do forwarding between addresses, so doesn't need to
address this issue.
You may be right about purely local forwarding, but what about
Charlie Brady wrote:
I didn't know there were plans. In any case, there is no need to detect
looping - the mail should simply be rejected before the data phase based
on the result of the MX lookup.
Loop detection is still necessary. Suppose you have two addresses that
forward to each other,
Hanno Hecker wrote:
we had some (smaller) troubles with a spammer sending mails with a
resolvable MX, but that MX pointed to 127.0.0.1... after blacklisting
these sending IP for a short time I added a check to the
require_resolvable_fromhost plugin...
I used to have my
Are there conditions under which the plugin should be returning DENY
rather than DENYSOFT? Sure if there's a timeout in the lookup then
DENYSOFT probably makes sense, but shouldn't NXDOMAIN, for example, rate
a DENY? Or is the idea that the sender will notice their mail queue and
fix their
I'm currently running qpsmtpd under 5.6.1, but I haven't updated qpsmtpd
in several weeks. I'd favor maintaining 5.6 compatibility, since there
are a lot of machines out there that came with 5.6 and haven't had their
perls upgraded. Not sure what the can't-do-without 5.8-isms are, though.
Bob Dodds wrote:
That doesn't help, so it must be the use of '=' instead
of a comma in $transaction-header-add(). perldoc
Mail::Header says to use commas there.
The = is equivalent to a comma in that context. And it has nothing to
do with whatever you're doing with at signs, either. The
Matt Sergeant wrote:
On 8 Jul 2005, at 22:20, Keith Ivey wrote:
The date line was missing but my mail server added one, so I didn't
notice.
Maybe that's something qpsmtpd should do too :-)
Hmm, I suppose I really shouldn't be subscribed to this list through a
mail server that I don't
I had another situation where a message was looping forever, being
repeatedly delivered from my server to itself, and causing a high load.
The problem is that the MX record looks like this:
conservation.com. 3600IN MX 1000 0.0.0.0.
I've previously encountered the same
John Peacock wrote:
You should upgrade to 0.29 at least (if
not the repository trunk), where the correct line would be
if ( $connection-relay_client() || exists($ENV{RELAYCLIENT}) )
Why doesn't -relay_client() check for $ENV{RELAYCLIENT}? Without that,
it seems you have to add ' ||
I'm going to YAPC in Toronto. Anybody else?
--
Keith C. Ivey [EMAIL PROTECTED]
Washington, DC
David Nicol wrote:
Hop counting is a fundamental part of reading the body -- as a plugin, it would
need to go in a per-body-line hook that is called for header lines, so the rest
of the message can be thrown out.
It might be more efficient that way (though as you say, the body of the
message
[EMAIL PROTECTED] wrote:
On Thu, 19 May 2005, Robert Spier wrote:
Either way, this should really be in a plugin.
I agree, but we should ship it with the distro, and have it enabled by
default in the example config.
What is the reason for making it a plugin? It's part of the basic
function of
Ask Bjørn Hansen wrote:
On May 18, 2005, at 9:20 PM, Keith Ivey wrote:
Any progress on this? I ran into a similar situation a while ago
when my server was trying to deliver a message to an address whose MX
record resolved to 0.0.0.0.
Why is it that qmail or postfix doesn't detect and stop
Keith Ivey wrote:
Without qpsmptd, this would be stopped by the loop-detection code in
qmail-smtpd, which is based on counting the Received headers. If
qpsmtpd is supposed to be a replacement, it should do that function of
qmail-smtpd.
Specifically, it appears (unless I'm missing something
John Peacock wrote:
Keith Ivey wrote:
I don't think there is any Delivered-To header yet, because the
message never gets delivered -- it's just repeatedly relayed (from the
server to itself). The hostname in the address is not in rcpthosts or
locals, so there's no delivery.
Almost exactly
On Thu, 19 May 2005, Juerd wrote:
By now I'm able to reproduce (not everything was going through this
server), and it happens only on large messages. The boundary is
somewhere between 8 and 10 kB.
Yes, I'm seeing the same thing. It's what was causing the duplicate
subject lines I mentioned
David Nicol wrote:
something like
if ((()=$Header =~ /^(Rec)/m) 100){ # ()= to create temp array to count
To replicate qmail-smtpd it would be more like
my $hops = 0;
$hops++ for $header-get('Received'), $header-get('Delivered-To');
if ( $hops = 100 ) {
I'm not sure why Received
John Peacock wrote:
I think this is what I should check in. Keith, can you confirm this
works for you???
Yes, that works fine. No loop when I temporarily removed the
config/logging and config/loglevel that I have now. I'm experimenting
with the adaptive logger. I'm still getting used to
John Peacock wrote:
Though a couple of self-inflicted Denial of Service attacks on my mail
servers, I've discovered that qpsmtpd doesn't implement any mailing loop
detection internally. Now that I have fixed my misconfiguration (alias
domains not completely set up), I thought I'd add this to
John Peacock wrote:
qmail-local.c: strerr_die1x(100,This message is looping: it
already has my Delivered-To line. (#5.4.6));
qmail-smtpd.c: if (hops) { out(554 too many hops, this message is
looping (#5.4.6)\r\n); return; }
But, I don't think that the qmail-local test is correct for
John Peacock wrote:
That would be caused by this line in Qpsmtpd::_load_plugins():
my ($plugin, @args) = split /\s+/, $plugin_line;
which splits the line on whitespace. I don't know if it is more
reasonable to say don't include leading whitespace in the config file
or just add a line like
Bob wrote:
What was the P for? perl? Queue Perl Daemon? Drop
smtp and that's qpd.
Qupid? That's been used for various things, but maybe not a
software project -- 579 Google hits isn't bad. No particularly
relevant connotations, but then Quench doesn't have that either.
--
Keith C. Ivey [EMAIL
Peter J. Holzer wrote:
On 2005-03-09 09:51:34 -0500, Keith Ivey wrote:
Qupius?
Pronounced almost but not entirely unlike copious? I like that.
Hmm, something you can fiddle with in your qupious free time.
Maybe that one wasn't so bad after all.
--
Keith C. Ivey [EMAIL PROTECTED]
Washington, DC
23 matches
Mail list logo
