Re: Problem with received from

2001-07-10 Thread Daniel Kelley


 The Received: from header has by 0 instead of by 
 starnie.harrison.org.  Looking in qmail-smtpd.c, I can see where it 
 gets assigned:

what do you have in /var/qmail/control/me?  that's one place to look.




Re: Problem with received from

2001-07-10 Thread Daniel Kelley



 
 Uh, looking back over my data, the tcpserver line has a 0 in the 
 location the tcpserver --help says should be a hostname.  Is this my 
 problem?  If so, I'd feel like a complete dumbass except for I just 
 checked and its like that in the LWQ text.

uh-oh:

from http://cr.yp.to/ucspi-tcp/tcpserver.html

Data-gathering options:
-l localname:   Do not look up the local host name in DNS; use localname for
the environment variable $TCPLOCALHOST. A common choice for localname 
is
0.

looks like that's what you want to change.  i'd try your test again, after
removing -l 0 from your qmail-smtpd/run script.

dan 




Re: qmailr processes

2001-07-10 Thread Daniel Kelley


 and restarted qmail, but I still get these processes sending me mail:
 
 qmailr  1071  0.0  0.5   888  568  ??  S12:27PM   0:00.01 qmail-remote
 graycastle.com query-return-28234-
 qmailr  1080  0.0  0.5   888  568  ??  S12:27PM   0:00.00 qmail-remote
 graycastle.com query-return-28236-
 ...

that's not incoming mail; qmail-remote send mail from your machine -
remote domains.

if this machine is supposed to be the primary MX for greycastle.com, you
have a comfiguration problem.

dan




Re: Unable_to_open_./Maildir:_is_a_directory._(#4.2.1)/

2001-07-05 Thread Daniel Kelley


you're probably starting qmail with and argument that looks like this:

qmail-start ./Maildir

instead of this:

qmail-start ./Maildir/

the trailing slash tells qmail that it should to a Maildir, and not
a file.  


 But when I try to send myself an email, it show line like this at my
 /var/log/qmail/current file :
 @40003b4528aa0eba06f4 starting delivery 4: msg 191485 to local
 [EMAIL PROTECTED]
 @40003b4528aa0eba7c24 status: local 1/10 remote 0/20
 @40003b4528aa1125658c delivery 4: deferral:
 Unable_to_open_./Maildir:_is_a_directory._(#4.2.1)/
 @40003b4528aa1125e674 status: local 0/10 remote 0/20





Re: Problem with VAR directory during install

2001-06-26 Thread Daniel Kelley



 I'm attempting to set up qmail on Mandrake 8.0. I do create 
 the /var/qmail directory, but when qmail compiles nothing goes 
 into this directory. Of course then when you run ./config it 
 barfs because it can't find certain things in /var/qmail. Yes, 
 I'm logged on as root. 

config does not fail because of anything in /var/qmail;  it fails if it
can't find your server's name in DNS.  Use config-fast instead.

 Also, when running the IDS script, I can't see that any 
 directories were created in the /var directory either. 

1. IDS is not a script that's included with qmail.
2. if you did create a IDS script off of INSTALL.ids (a la LWQ), it
doesn't create any directories anywhere.  It adds user accounts and groups
to your system.

 I've reloaded Mandrake 3 times from scratch and it's still doing 
 the same thing. Also tried gcc instead of cc and it makes no 
 difference. 

none of these things have anything to do with installing qmail.

 What am I doing wrong? 

you need to read and understand the various INSTALL. files included with
your qmail distribution.

 BTW, the /var directory resides on a separate physical drive in 
 the machine - a Linux RAID 1 (mirror), to be precise, so there 
 are two drives in fact. The /var directory shows up nicely and 
 is perfectly accessible. I can't imagine this would be creating 
 the problem. 

It's not.  You need to better understand the qmail install process. start
by reading the documentation included with qmail, then visit
lifewithqmail.org.

Dan




Re: qmail-injecting a message with 50K Bcc:

2001-06-25 Thread Daniel Kelley


 Wrong. Ezmlm is what you need. It's a high speed mailing list manager,
 and with the qmail-verh patch you can have individual addressing. You
 can also take input from a text file of one address per line when
 subscribing the list members.

does this hold true for one-time mailings?  i'm sending a very dry email
detailing the ownership change of a corportaion, so i can't forsee many
responses (bounces are, of course, another story).

the reason that i originally tried to do this with qmail-inject instead of
elmlm was that i never saw a need to have list-like behavior (replies,
postings,etc).  that being tha case, is ezmlm still the best option?

thanks-

dan 





reasearch: qmail-qfilter freebsd segfault

2001-06-08 Thread Daniel Kelley


hi all-

with some pointers from charles cazabon, i've stared to trace the
occurance of qmail-qfilter segfaults on freebsd.

here's what i've done:

i set QMAILQUEUE to /var/qmail/bin/qmail-runfilters via tcpserver for all
incoming mail:

/etc/tcp.smtp:
:allow,QMAILQUEUE=/var/qmail/bin/qmail-runfilters 

i'm tracing the execution of every instance of qmail-qfilter via the
/var/qmail/bin/qmail-runfilters script:

#!/bin/sh
exec ktrace -i -f /tmp/ktrace.output`date +%H%M%S` \
/var/qmail/bin/qmail-qfilter /var/qmail/qfilter/filters/deny-filetypes

/var/qmail/qfilter/filters/deny-filetypes is taken directly from the
samples director that comes with qmail-qfilter-1.4.

here's a sample kdump of a trace file generated by a mail that causes
qmail-qfilter to segfault:

58293 qmail-qfilter RET   write 1024/0x400
 58293 qmail-qfilter CALL  read(0,0xbfbfec68,0x1000)
 58293 qmail-qfilter GIO   fd 0 read 0 bytes
   
 58293 qmail-qfilter RET   read 0
 58293 qmail-qfilter CALL  lseek(0x5,0,0,0,0)
 58293 qmail-qfilter RET   lseek 0
 58293 qmail-qfilter CALL  read(0x1,0xbfbfec68,0x1000)
 58293 qmail-qfilter GIO   fd 1 read 0 bytes
   
 58293 qmail-qfilter RET   read 0
 58293 qmail-qfilter PSIG  SIGSEGV SIG_DFL
 58293 qmail-qfilter NAMI  qmail-qfilter.core

i't looks like the segfault occurs after qmail-qfilter has created the
temp file to hold the message, when it's trying to read the message back.

in an attempt to isolate anything in particular emails that would cause
the problem, i uncommented the unlink() call in qmail-qfilter.c that makes
the temp files vanish.  

i'm attaching two text files to this email:
1. the complete temp file created bu qmail-qfilter
2. the complete trace file generated by kdumping the ktrace file generated
when qmail-qfilter segfaults.

if anyone has any comments on any of this, or any ideas on why the
read() syscall might cause the segfault, it'd be much appreciated.  

thanks-

dan




Re: qmail ONLY selectively receiving mail from outside

2001-06-04 Thread Daniel Kelley


what you want to do is allow all incoming connections (i.e. don't have
your /etc/hosts.allow  /etc/hosts.deny setup to drop all miscellaneous
smtp connections) on port 25.  if you only set RELAYCLIENT for the ip's
you want, qmail will handle rejecing the emails.

make sense?

On Mon, 4 Jun 2001, Ashe Coutts wrote:

 smallerI have set up a qmail system (RedHat linux 7.1, kernel 2.4.5, xinetd, qmail 
 1.03 RPMs, U of Wash pop3 and imap, etc.). with a domain name of 
 sbcacademy.org (machine name mail.sbcacademy.org)  with the 
 following configuration files:
 
 
 === start  /var/qmail/control/rcpthosts 
 
 localhost
 
 sbcacademy.org
 
 mail.sbcacademy.org
 
 === end  /var/qmail/control/rcpthosts 
 
 
 === start excerpt from /etc/hosts.allow 
 
 ### The qmail outgoing/retrieval stuff
 
 ipop3d, imapd  :  ALL
 
 
 FontFamilyparamCourier New/param### The qmail selective relaying stuff
 
 tcp-env  :  xx.yy.zz.:  setenv RELAYCLIENT
 
 tcp-env  :  aa.bb.cc.dd  :  setenv RELAYCLIENT
 
 tcp-env  :  localhost  :  setenv RELAYCLIENT 
 
 === end excerpt from /etc/hosts.allow FontFamilyparamArial/param
 
 
 The system works but will not receive mail from outside the select few I
 
 list in the /etc/hosts.allow file.
 
 
 I think I now understand what is going on but not why. 
 
 
 I can appreciate that relaying is a bad idea but fail to see how to 
 
 set up the qmail so anyone can at least send mail to a user on 
 
 the qmail system.  Receiving from anywhere would be our problem 
 
 and would not be relaying anything beyond us.
 
 
 As is, ONLY mail coming from our system (xx.yy.zz.) or our county 
 
 education email server (aa.bb.cc.dd) is received.
 
 
 In testing from an outside hotmail account I was seeing the following in
 
 /var/log/messages:
 
 
 FontFamilyparamCourier New/paramsmaller=== Start excerpt from 
/var/log/messages =
 
 Jun  4 09:13:19 mail xinetd[492]: refused connect from 
 
 209.185.241.98
 
 Jun  4 09:19:57 mail xinetd[492]: refused connect from 
 
 209.185.241.80
 
 === End from /var/log/messages =
 
 
 FontFamilyparamArial/parambiggerSo I entered this next line in 
/etc/hosts.allow:
 
 tcp-env  :  209.185.241.   :  setenv RELAYCLIENT
 
 and viola - in comes a message from the hotmail account to one of 
 
 our users.
 
 
 I want qmail to accept email from other email users/systems in the world
 
 as other email systems I've set up do WITHOUT having to explicitly enter
 
 every email system I want to receive mail from.
 
 
 Can qmail allow for this and still prevent relay abuse or are the two 
 somehow tied together??
 
 
 I have read what I could on the lists regarding what I'm trying to 
 accomplish (FAQ, The qmail newbie's guide to relaying, Selective 
 relaing with tcpserver and qmail-smtpd, etc.) and realize that this topic is 
 almost a dead horse BUT I'm still unable to understand what is required to 
 do what I wish so thanks in advance to any suggestions and/or 
 recommendations any of you may offer.
 
 
 nofill
 Experience is not what happens to you, it
  is what you do with what happens to you
-- Aldous Huxley (1894-1963)
 
Ashe Coutts ([EMAIL PROTECTED])
805.963.4338 Ext 300
Fax 805.884.1557
 




qmail-qfilter signal 11

2001-06-04 Thread Daniel Kelley


hi-

i think that someone posted earlier today regarding sporadic sig 11's on
freebsd 4.2-RELEASE while running qmail-qfilter.

interestingly, i just installed qmail-qfilter earlier today on the same
release of freebsd, and i'm getting the same thing:

Jun  4 19:46:03 mx1 /kernel: pid 64541 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 19:48:39 mx1 /kernel: pid 64670 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 20:01:07 mx1 /kernel: pid 65370 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 20:04:58 mx1 /kernel: pid 65495 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 20:18:38 mx1 /kernel: pid 66276 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 20:50:45 mx1 /kernel: pid 68131 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 21:02:07 mx1 /kernel: pid 68625 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 21:52:43 mx1 /kernel: pid 70813 (qmail-qfilter), uid 1003: exited
on signal 11
Jun  4 22:26:45 mx1 /kernel: pid 71658 (qmail-qfilter), uid 1003: exited
on signal 11

the previous thread on this topic ended when it was suggested that the
problem was running softlmit with a limit of 2 meg, which didn't give the
perl interpreter enough room to get started.  

interestingly enough, the problem persists after removing the softlimit
from /var/qmail/supervise/qmail-smtpd/run. (i'm running LWQ-style)

i've checked the sig 11 FAQ (www.bitwizard.nl/sig11), but that doesn't
seem to have anything relevant.

has anyone else running qmail on freebsd noticed this behavior?

thanks-

dan






Re: Re[2]: Oops,I guess Sendmail wasn't secure after all...

2001-06-02 Thread Daniel Kelley


 Why no one makes a package with all you need to download and
 install, here is a suggestion:
 
 - qmail
 - the tcpserver
 - something good for pop before smtp
 - vpopmail
 - good tools for blocking spam, blocking mails from open relays, and
 so on
 - and other additions from other people i do not know
 
 There should be one file to download and the makefile should do nearly
 everything neccessary. I should not spend days to understand the
 different modules as a newbie, it takes too much time.

the author of qmail has specific rules for how qmail packages can be
distributed. 

see http://cr.yp.to/qmail/dist.html

basically, you can distribute so called var-qmail packages, but anything
else seems to require the Dan Bernstein's approval.




Re: Relay for domains in rcpthosts

2001-04-30 Thread Daniel Kelley


i think you're essentially referring to how to maintain a secondary MX.

qmail will not attempt local delivery to any host that is in rcpthosts,
but is not in locals/me.

if you are running under tcpserver, make sure that you're not preventing
any connections you want, and you should be fine.


On Mon, 30 Apr 2001 [EMAIL PROTECTED] wrote:

 I want to allow relaying of all domains in my rcpthosts file. While I
 know this is a very bad thing to do this is only a very temporary measure.
 
 I have searched high and low for this...and I cant find the answer!
 
 Thanks,
 
 ross
 
 
 




Re: TCPserver; ucspi-tcp; inetd

2001-04-26 Thread Daniel Kelley


you already have a service listening on port 25.  a likely culprit is
sendmail.  make sure that it's shut off, and make sure that you check
/etc/inetd.conf.  if there's something in there listening on the smtp
port, comment it out, and kill -HUP inetd's pid.

dan

On Fri, 27 Apr 2001, jx001 wrote:

 When I did  install the ucspi-tcp package I receive the following
 error-message (var/log/qmail/qmail-smtpd):
 tcpserver: fatal: unable to bind: adress already use
 
 I did install ucspi-tcp as described (http://cr.yp.to/ucspi-tcp.html)
 
 There must be some problems with inetd...
 
 How can I solve this problem?
 
 THX
 
 Joe
 




Re: Should I worry about these compile warnings?

2001-03-29 Thread Daniel Kelley


no.

On Thu, 29 Mar 2001, Sid Wilroy wrote:

 maildirwatch.c:62: warning: return type of `main' is not `int'
 ./load maildirwatch hfield.o headerbody.o maildir.o \
 prioq.o now.o getln.a env.a open.a strerr.a stralloc.a \
 alloc.a substdio.a error.a str.a
 
 idedit.c:124: warning: return type of `main' is not `int'
 ./load idedit strerr.a substdio.a error.a str.a fs.a \
 wait.a open.a seek.a
 
 install.c:156: warning: return type of `main' is not `int'
 ./load install-big fifo.o install.o auto_qmail.o \
 auto_split.o auto_uids.o strerr.a substdio.a open.a error.a \
 str.a fs.a
 
 
 




Re: Repeated Identical Messages

2001-03-17 Thread Daniel Kelley


i think the latter is correct.  this is my fault;  i'm responsible for
setting up our secondary, and the admin who set up our primary told me
that it was qmail-1.03.  it's definitely qmail, but must be an older
version.

thanks-
dan



On Sat, 17 Mar 2001, Peter van Dijk wrote:

 On Fri, Mar 16, 2001 at 12:52:40AM +0100, Markus Stumpf wrote:
  On Thu, Mar 15, 2001 at 05:05:30PM -0500, dan kelley wrote:
   they're definitley qmail;  both run qmail-1.03 unpatched under tcpserver.
   
   what do you mean by 'help for smtp' ?
  
  
  lagrange(2:2697) $ telnet mailhost.otec.com smtp
  Trying 209.3.117.5...
  Connected to mx1.ny.otec.com.
  Escape character is '^]'.
  220 *
  help
  502 unimplemented (#5.5.1)
  quit
  221 mx1.ny.otec.com
 
 It does give me the 'timestamp qp pid' thing when I send a message
 thru it... Looks like heavily patched, or maybe old qmail.
 
 Greetz, Peter.
 




recordio / logging problem

2001-03-11 Thread Daniel Kelley


hi-

i'm trying to record all incoming smtp sessions using recordio, but i'm
stymied by the following:

/usr/local/bin/tcpserver -R -x /etc/tcp.smtp.cdb -c 100 -u user -g
group 0 smtp \
/usr/local/bin/rblsmtpd /usr/local/bin/recordio /var/qmail/bin/qmail-smtpd
\
| /var/qmail/bin/splogger smtpd 3 

this sends all recordio output to the terminal, not through syslog.
following a couple of examples on this list, i inserted 21 directly
after qmail-smtpd, but that generates 'Ambiguous Ouptut Redirect'.

This is probably just a stupid shell thing (i'm using csh on linux), but i
don't understand it.

thanks-

dan





451 timeout

2001-03-11 Thread Daniel Kelley


Hi all-

I've been expereincing a lot of ugly errors lately:  many copies of the
smae emial arriving, some remte hosts can't email us at all (yes, MX
records are ok).  I've begun to watch everytihng that happens w/ recordio.  

In my log files, I'm beginning to see a distressing number of errors like
the one listed below:

a remote server connects, announces MAIL, RCPT TO, and DATA fine, then the
connection is dropped and a '451 error' code is logged.  In the example
below, this happened only 22 seconds after the previous entry.  This is
pretty representative of what's happening.

As I understand it, 451 is an internal error while processing a message,
and the default timeout for smtp sessions (timeoutsmtpd) is 1200 seconds.

Has anyone seen anything like this, and if so, could it be a configuration
error of some sort?  Or does it indicate either hardware or network
failure?

Thanks-

Dan


Mar 11 19:38:23 mx1 smtpd: 984357503.141262 13096  220 mx1.ny.otec.com
ESMTP?
Mar 11 19:38:23 mx1 smtpd: 984357503.216899 13096  HELO
femail9.sdc1.sfba.home.com?
Mar 11 19:38:23 mx1 smtpd: 984357503.218213 13096  250 mx1.ny.otec.com?
Mar 11 19:38:23 mx1 smtpd: 984357503.414713 13096  MAIL
FROM:[EMAIL PROTECTED]?
Mar 11 19:38:23 mx1 smtpd: 984357503.416019 13096  250 ok?
Mar 11 19:38:23 mx1 smtpd: 984357503.490315 13096  RCPT
TO:[EMAIL PROTECTED]?
Mar 11 19:38:23 mx1 smtpd: 984357503.491628 13096  250 ok?
Mar 11 19:38:23 mx1 smtpd: 984357503.566791 13096  DATA?
Mar 11 19:38:23 mx1 smtpd: 984357503.576821 13096  354 go ahead?
Mar 11 19:38:45 mx1 smtpd: 984357525.672773 12927  451 timeout (#4.4.2)?
Mar 11 19:38:45 mx1 smtpd: 984357525.673019 12927  [EOF]






Re: Still having mutiple delivery problems.

2001-03-07 Thread Daniel Kelley


i am having the *exact* same problem.  it only happens with mails
originating outside our domain, and tehre's no rhyme or reason to it.  
our logs do show 

seems to happen more often with hotmail and yahoo.  it's also happend with
this list!  the only constant i can see between the three is that they all
(i think) use qmail for outgoing mail.  

i can't comfirm that this has happend at all when the mail server is
anything but qmail.  how could this be the common link?

dan

On Wed, 7 Mar 2001, Kep Brown wrote:

 Ok I am still having problems with the same emails getting delivered
 multiple times. And am baffled by what to do.  The higher ups are pushing
 REAL hard to switch to Microsoft Exchange so I need at least some ideas on
 what to try next.
 
 History,
Certain email messages, sent from outside our local network, with the
 Reply to header set to our home domain, ie. [EMAIL PROTECTED] sent using local
 ISPs email server, get delivered multiple times.  It is not consistant, one
 time a message sent to certain users will only get delivered once, the next
 time a message sent to the same users will start looping, get delivered
 multiple times.
 
 Information
Far servers logs show a lost connection while sending end of data.
local sever shows no errors in maillog or messages.  With recordio turned
 on the conversations look the same for correctly delivered messages as
 incorrectly delivered messages.
 
 ANY help or suggestions or.. would be greatly appreciated.
 
 Thanks
 Kep
 
 
 
 
 
  Kep Brown
   Systems, Network and Database Administrator
   phone: (805) 560-3781
   fax: (805) 560-3991
   e: [EMAIL PROTECTED]
 
 
 
 
 
 




Re: Still having mutiple delivery problems.

2001-03-07 Thread Daniel Kelley


aha!  we use cucipop as well.  there's only one 'error locking' message in
the logs, but there are plenty of these:

cucipop[20914]: Invalid command capa - -

i've tried to figure out what this means, but i have no clue.  could this
be indicative of what you're talking about?

if outlook clients are sending messages that cause this, what steps can be
taken to prevent it?

thanks-

dan


On Wed, 7 Mar 2001 [EMAIL PROTECTED] wrote:

 On Wed, Mar 07, 2001 at 03:03:20PM -0500, Daniel Kelley wrote:
  
  i am having the *exact* same problem.  it only happens with mails
  originating outside our domain, and tehre's no rhyme or reason to it.  
  our logs do show 
  
  seems to happen more often with hotmail and yahoo.  it's also happend with
  this list!  the only constant i can see between the three is that they all
  (i think) use qmail for outgoing mail.  
  
  i can't comfirm that this has happend at all when the mail server is
  anything but qmail.  how could this be the common link?
 
 Are you seeing this as SMTP level?  We get duplicated messages frequently
 but it's always (as far as I can tell) a pop issue.  It is always solved by
 deleting an improperly formatted multipart mime message, almost always
 from outlook.  The logs show somthing like "unable to lock mailbox; connection
 lost" or something similar.  That's with cucipop.
 
  
  dan
  
  On Wed, 7 Mar 2001, Kep Brown wrote:
  
   Ok I am still having problems with the same emails getting delivered
   multiple times. And am baffled by what to do.  The higher ups are pushing
   REAL hard to switch to Microsoft Exchange so I need at least some ideas on
   what to try next.
   
   History,
  Certain email messages, sent from outside our local network, with the
   Reply to header set to our home domain, ie. [EMAIL PROTECTED] sent using local
   ISPs email server, get delivered multiple times.  It is not consistant, one
   time a message sent to certain users will only get delivered once, the next
   time a message sent to the same users will start looping, get delivered
   multiple times.
   
   Information
  Far servers logs show a lost connection while sending end of data.
  local sever shows no errors in maillog or messages.  With recordio turned
   on the conversations look the same for correctly delivered messages as
   incorrectly delivered messages.
   
   ANY help or suggestions or.. would be greatly appreciated.
   
   Thanks
   Kep
   
   
   
   
   
Kep Brown
 Systems, Network and Database Administrator
 phone: (805) 560-3781
 fax: (805) 560-3991
 e: [EMAIL PROTECTED]
   
   
   
   
   
   
 
 -- 
 
 Christopher F. Miller, Publisher [EMAIL PROTECTED]
 MaineStreet Communications, Inc 208 Portland Road, Gray, ME  04039
 1.207.657.5078   http://www.maine.com/
 Content management, electronic commerce, internet integration, Debian linux
 




Re: Still having mutiple delivery problems.

2001-03-07 Thread Daniel Kelley



On Wed, 7 Mar 2001 [EMAIL PROTECTED] wrote:

 On Wed, Mar 07, 2001 at 04:00:29PM -0500, Daniel Kelley wrote:
  
  aha!  we use cucipop as well.  there's only one 'error locking' message in
  the logs, but there are plenty of these:
  
  cucipop[20914]: Invalid command capa - -
  
  i've tried to figure out what this means, but i have no clue.  could this
  be indicative of what you're talking about?
  
  if outlook clients are sending messages that cause this, what steps can be
  taken to prevent it?
  
  thanks-
  
  dan
 
 It looks like this:
 Mar  5 18:02:29 gray cucipop[15401]: Error locking hwware's mailbox 
 Mar  5 18:02:29 gray cucipop[15401]: lost hwware 208.148.249.48 240, 16 (752895), 0 
(0)
 
 Note that this is a client issue, not qmail.  You need to find out
 what "getting delivered multiple times" **really** means.  Get the headers.
 Anytime anyone uses passive tense to describe what is going on - red flags.

ok, i'm now pretty sure that this is a qmail/SMTP issue:

in examining the mboxes of those affected, they each have many copies of
the duplicate emails, so it's not the case that cucipop is simply
making it appear as if they have received the message 100 times;  there
are actually 100 copies of the mail in the box.

the date sent is identical on every message, but the message is actually
being received many times by qmail on our end.  

here are two examples of the exact same message being delivered twice
(from this mailing list).  i know that qmail tries to verify delivery, so
some situation must exist where my qmail installation is not providing
that confirmation to the remote mailserver.  

here are the messages:

message 1:

From [EMAIL PROTECTED] Wed Mar 07 22:04:21
2001
Return-Path: [EMAIL PROTECTED]
Delivered-To: [EMAIL PROTECTED]
Received: (qmail 20442 invoked from network); 7 Mar 2001 22:04:20 -
Received: from muncher.math.uic.edu (131.193.178.181)
  by mx1.ny.otec.com with SMTP; 7 Mar 2001 22:04:20 -
Received: (qmail 23799 invoked by uid 1002); 1 Mar 2001 19:13:31 -
Mailing-List: contact [EMAIL PROTECTED]; run by ezmlm
Precedence: bulk
Delivered-To: mailing list [EMAIL PROTECTED]
Received: (qmail 24628 invoked from network); 1 Mar 2001 19:13:31 -
Received: from mat.tgd.net (HELO man.tgd.net) (64.81.67.116)
  by muncher.math.uic.edu with SMTP; 1 Mar 2001 19:13:31 -
Received: (qmail 82115 invoked by uid 1000); 1 Mar 2001 19:12:41 -
Date: Thu, 1 Mar 2001 11:12:41 -0800
From: Sean Chittenden [EMAIL PROTECTED]
To: "Rob Hines Jr." [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Subject: Re: Scalable Mail Solution
Message-ID: [EMAIL PROTECTED]
References: [EMAIL PROTECTED]
[EMAIL PROTECTED]
Mime-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha1;
protocol="application/pgp-signature"; boundary="0DcgAuOrObJpcAxl"
Content-Disposition: inline
User-Agent: Mutt/1.2.5i
 

message 2:

Return-Path: [EMAIL PROTECTED]
Delivered-To: [EMAIL PROTECTED]
Received: (qmail 11264 invoked from network); 2 Mar 2001 01:12:55 -
Received: from muncher.math.uic.edu (131.193.178.181)
  by mx1.ny.otec.com with SMTP; 2 Mar 2001 01:12:55 -
Received: (qmail 23799 invoked by uid 1002); 1 Mar 2001 19:13:31 -
Mailing-List: contact [EMAIL PROTECTED]; run by ezmlm
Precedence: bulk
Delivered-To: mailing list [EMAIL PROTECTED]
Received: (qmail 24628 invoked from network); 1 Mar 2001 19:13:31 -
Received: from mat.tgd.net (HELO man.tgd.net) (64.81.67.116)
  by muncher.math.uic.edu with SMTP; 1 Mar 2001 19:13:31 -
Received: (qmail 82115 invoked by uid 1000); 1 Mar 2001 19:12:41 -
Date: Thu, 1 Mar 2001 11:12:41 -0800
From: Sean Chittenden [EMAIL PROTECTED]
To: "Rob Hines Jr." [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Subject: Re: Scalable Mail Solution
Message-ID: [EMAIL PROTECTED]
References: [EMAIL PROTECTED]
[EMAIL PROTECTED]
Mime-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha1;
protocol="application/pgp-signature"; boundary="0DcgAuOrObJpcAxl"
Content-Disposition: inline
User-Agent: Mutt/1.2.5i



 
  On Wed, 7 Mar 2001 [EMAIL PROTECTED] wrote:
  
   On Wed, Mar 07, 2001 at 03:03:20PM -0500, Daniel Kelley wrote:

i am having the *exact* same problem.  it only happens with mails
originating outside our domain, and tehre's no rhyme or reason to it.  
our logs do show 

seems to happen more often with hotmail and yahoo.  it's also happend with
this list!  the only constant i can see between the three is that they all
(i think) use qmail for outgoing mail.  

i can't comfirm that this has happend at all when the mail server is
anything but qmail.  how could this be the common link?
   
   Are you seeing this as SMTP level?  We get duplicated messages frequently
   but it's always (as far as I can tell) a pop issue.  It is always solved by
   deleting an improperly formatted multipart mime message, almost always
   from outlook.  T

QMTP/mail distribution

2001-03-04 Thread Daniel Kelley


hi all-

i currently have a setup where one central mail server receives all mail
for a primary domain, then redistributes it to subdomains (eg satellite
offices) as necessary.  this is currently done with fastforward and SMTP.
i'd like to try out QMTP, and use this to forward mail from the central
server out to the satellites.  is there an easy way to cofigure the cenral
server to know who it can/should speak QMTP with?  it seems very
straightforward to receive QMTP messages, but i'm not sure if there's a
great way to decide when to send QMTP mail.

thanks-

dan