qmail Digest 18 Jun 1999 10:00:00 -0000 Issue 675
Topics (messages 26803 through 26866):
ETRN
26803 by: Anand Buddhdev <[EMAIL PROTECTED]>
26864 by: Anand Buddhdev <[EMAIL PROTECTED]>
move the queue
26804 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
26809 by: Van Liedekerke Franky <[EMAIL PROTECTED]>
qmail on Solaris 2.7
26805 by: "Mark E. Drummond" <[EMAIL PROTECTED]>
TEST.receive failure
26806 by: "Alex Miller" <[EMAIL PROTECTED]>
26807 by: "Alex Miller" <[EMAIL PROTECTED]>
Multiple qMail Instances
26808 by: Peter Haworth <[EMAIL PROTECTED]>
TEST ONLY, excuse me.
26810 by: Georgi Kupenov <[EMAIL PROTECTED]>
Tower of Babel problem
26811 by: "Alex Miller" <[EMAIL PROTECTED]>
26813 by: "Paul J. Schinder" <[EMAIL PROTECTED]>
26814 by: John O'Shea <[EMAIL PROTECTED]>
26815 by: "Alex Miller" <[EMAIL PROTECTED]>
26816 by: Paul Schinder <[EMAIL PROTECTED]>
26817 by: "Alex at Star" <[EMAIL PROTECTED]>
26844 by: "Chris Garrigues" <[EMAIL PROTECTED]>
blocking trash from "Internet Mail Service (5.5.xxxx.x)
26812 by: "Adam D. McKenna" <[EMAIL PROTECTED]>
26839 by: torben fjerdingstad <[EMAIL PROTECTED]>
/bin/mail
26818 by: "Matt Schnierle" <[EMAIL PROTECTED]>
Closing SMTP connection
26819 by:
26822 by: Balazs Nagy <[EMAIL PROTECTED]>
26823 by: Justin Bell <[EMAIL PROTECTED]>
26824 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
26826 by: Balazs Nagy <[EMAIL PROTECTED]>
26827 by: "Alex Miller" <[EMAIL PROTECTED]>
26828 by: Stefan Paletta <[EMAIL PROTECTED]>
26829 by: Justin Bell <[EMAIL PROTECTED]>
26833 by: Scott Schwartz <[EMAIL PROTECTED]>
26834 by:
26835 by: Justin Bell <[EMAIL PROTECTED]>
26840 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
26847 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
syslog and load avg at 11,000 msgs per day
26820 by: Eric Dahnke <[EMAIL PROTECTED]>
26821 by: Balazs Nagy <[EMAIL PROTECTED]>
26825 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
26838 by: Richard Letts <[EMAIL PROTECTED]>
26841 by: "Matt Schnierle" <[EMAIL PROTECTED]>
relaying question
26830 by: Jeffrey Finkelstein <[EMAIL PROTECTED]>
26837 by: Stefan Paletta <[EMAIL PROTECTED]>
26842 by: Jeffrey Finkelstein <[EMAIL PROTECTED]>
26845 by: Stefan Paletta <[EMAIL PROTECTED]>
26856 by: "Timothy L. Mayo" <[EMAIL PROTECTED]>
Loop ?
26831 by: "Alberto Schiappa" <[EMAIL PROTECTED]>
26866 by: Giles Lean <[EMAIL PROTECTED]>
cron complaing about qmail
26832 by: Giles Lean <[EMAIL PROTECTED]>
26843 by: "Chris Garrigues" <[EMAIL PROTECTED]>
tool for verifying addresses
26836 by: [EMAIL PROTECTED]
cyclog examples sought
26846 by: John R Levine <[EMAIL PROTECTED]>
26860 by: Russell Nelson <[EMAIL PROTECTED]>
26861 by: Anand Buddhdev <[EMAIL PROTECTED]>
Killing qmail - is it safe?
26848 by: [EMAIL PROTECTED]
26851 by: Richard Letts <[EMAIL PROTECTED]>
26855 by: Dave Kitabjian <[EMAIL PROTECTED]>
26857 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
Cyrus question re access/sharing
26849 by: Eric Shafto <[EMAIL PROTECTED]>
26853 by: Dave Teske <[EMAIL PROTECTED]>
Boot file
26850 by: James <[EMAIL PROTECTED]>
26852 by: "Soffen, Matthew" <[EMAIL PROTECTED]>
26854 by: "Alex Miller" <[EMAIL PROTECTED]>
large attachments, tcpserver & timeouts
26858 by: Aaron Nabil <[EMAIL PROTECTED]>
26859 by: "Sam" <[EMAIL PROTECTED]>
26863 by: Aaron Nabil <[EMAIL PROTECTED]>
26865 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
[help!] adding trailer on each outgoing message?
26862 by: Silver CHEN <[EMAIL PROTECTED]>
Administrivia:
To subscribe to the digest, e-mail:
[EMAIL PROTECTED]
To unsubscribe from the digest, e-mail:
[EMAIL PROTECTED]
To bug my human owner, e-mail:
[EMAIL PROTECTED]
To post to the list, e-mail:
[EMAIL PROTECTED]
----------------------------------------------------------------------
On Thu, Jun 17, 1999 at 12:06:11PM +0545, HTP Communication Pvt. Ltd wrote:
*smile* Exchange is trying to be clever by parsing the output of its EHLO
to qmail-smtpd! That's a first.
Anyway, qmail *does* support ETRN - it just doesn't advertise it. You'll
need to do 2 things: First, download serialmail-0.75, install it, read the
AUTOTURN file and setup the necessary configuration for your customer. Your
current setup doesn't sound right.
Secondly, you'll have to patch qmail-smtpd.c to make it advertise ETRN.
It's a very simply patch. Look for the following code in qmail-smtpd.c:
void smtp_ehlo(arg) char *arg;
{
smtp_greet("250-"); out("\r\n250-PIPELINING\r\n250 8BITMIME\r\n");
seenmail = 0; dohelo(arg);
}
And change it to:
void smtp_ehlo(arg) char *arg;
{
smtp_greet("250-");
out("\r\n250-PIPELINING\r\n250-8BITMIME\r\n250 ETRN\r\n");
seenmail = 0; dohelo(arg);
}
And then recompile with "make". Finally, copy the new qmail-smtpd binary
from this source directory to your qmail bin directory.
Maybe DJB can add this to his TODO list for qmail 2.0. I've also been
contemplating doing something about the SIZE parameter of ESMTP. qmail
does support a size limit for incoming mail, but doesn't advertise it - as
a consequence, big messages end up being transmitted, before they are
rejected. Thing is, I'm not good with C programming, and don't feel
comfortable with making major code changes. DJB: Any comments?
> Hello,
> I have a client who is using M/S Exchange 5.5 to download/upload the emails
> from my server. I have qmail-1.03 running in my server. He has a domain
> called blah.htp.com.np and is give a fixed IP address from my server. I have
> an MX entry for his domain as well and created a virtual doamain such that
> all the messaes for his domains comes to his mail box. But whenever he tries
> to send the mails from my server it says
>
> Internet mali service is configured to used ETRN to Dequeue mail but
> 206.82.132.2 does not support ETRN
>
> what does this means, does not qmail supports ETRN?
>
> I am not on the list, pls reply personally.
--
Anand
On Thu, Jun 17, 1999 at 01:05:07PM +0300, Anand Buddhdev wrote:
A follow on to my own post. In my hurry to go have lunch, I failed to note
that simply advertising ETRN isn't enough. qmail-smtpd.c must also
acknowledge an ETRN request with at least a "250 ok". So I added a bit more
code to do that. I apologise to everyone for my mistake.
I added an smtp_etrn subroutine after the smtp_help subroutine:
..
..
void smtp_help()
{
out("214 qmail home page: http://pobox.com/~djb/qmail.html\r\n");
}
void smtp_etrn()
{
out("250 ok\r\n");
}
void smtp_quit()
{
smtp_greet("221 "); out("\r\n"); flush(); _exit(0);
}
..
..
And then the following line in the SMTP commands structure:
struct commands smtpcommands[] = {
{ "rcpt", smtp_rcpt, 0 }
, { "mail", smtp_mail, 0 }
, { "data", smtp_data, flush }
, { "quit", smtp_quit, flush }
, { "helo", smtp_helo, flush }
, { "ehlo", smtp_ehlo, flush }
, { "rset", smtp_rset, 0 }
, { "help", smtp_help, flush }
, { "noop", err_noop, flush }
, { "vrfy", err_vrfy, flush }
, { "etrn", smtp_etrn, flush } <----- inserted line
, { 0, err_unimpl, flush }
} ;
I think this is better. Although the "250 ok" in response to ETRN is not
really needed (I could care less what the client machine thinks of the
response), it is the right thing to do.
> On Thu, Jun 17, 1999 at 12:06:11PM +0545, HTP Communication Pvt. Ltd wrote:
>
> *smile* Exchange is trying to be clever by parsing the output of its EHLO
> to qmail-smtpd! That's a first.
>
> Anyway, qmail *does* support ETRN - it just doesn't advertise it. You'll
> need to do 2 things: First, download serialmail-0.75, install it, read the
> AUTOTURN file and setup the necessary configuration for your customer. Your
> current setup doesn't sound right.
>
> Secondly, you'll have to patch qmail-smtpd.c to make it advertise ETRN.
> It's a very simply patch. Look for the following code in qmail-smtpd.c:
>
> void smtp_ehlo(arg) char *arg;
> {
> smtp_greet("250-"); out("\r\n250-PIPELINING\r\n250 8BITMIME\r\n");
> seenmail = 0; dohelo(arg);
> }
>
> And change it to:
>
> void smtp_ehlo(arg) char *arg;
> {
> smtp_greet("250-");
> out("\r\n250-PIPELINING\r\n250-8BITMIME\r\n250 ETRN\r\n");
> seenmail = 0; dohelo(arg);
> }
>
> And then recompile with "make". Finally, copy the new qmail-smtpd binary
> from this source directory to your qmail bin directory.
>
> Maybe DJB can add this to his TODO list for qmail 2.0. I've also been
> contemplating doing something about the SIZE parameter of ESMTP. qmail
> does support a size limit for incoming mail, but doesn't advertise it - as
> a consequence, big messages end up being transmitted, before they are
> rejected. Thing is, I'm not good with C programming, and don't feel
> comfortable with making major code changes. DJB: Any comments?
>
> > Hello,
> > I have a client who is using M/S Exchange 5.5 to download/upload the emails
> > from my server. I have qmail-1.03 running in my server. He has a domain
> > called blah.htp.com.np and is give a fixed IP address from my server. I have
> > an MX entry for his domain as well and created a virtual doamain such that
> > all the messaes for his domains comes to his mail box. But whenever he tries
> > to send the mails from my server it says
> >
> > Internet mali service is configured to used ETRN to Dequeue mail but
> > 206.82.132.2 does not support ETRN
> >
> > what does this means, does not qmail supports ETRN?
> >
> > I am not on the list, pls reply personally.
>
> --
> Anand
--
System Administrator
See complete headers for address, homepage and phone numbers
+ Van Liedekerke Franky <[EMAIL PROTECTED]>:
| 929610000.245483 alert: cannot start: hath the daemon spawn no fire?
As far as I can tell from a quick look at the source, that means that
either qmail-lspawn or qmail-rspawn fails to communicate. You might
wish to trace system calls and see what is going on.
| I followed the hints for moving the
| queue like told in the README file provided with queue-fix.
| Is it possible that qmail doesn't like a symbolic link?
Not likely. qmail-send just changes to the queue directory, and
hollers if it can't (a different kind of hollering than you saw).
Actually, now that I think of it, I would suspect that something bad
happened to the files in queue/lock when you moved the directory to a
different file system. I am not sure if queue-fix fixes that?
- Harald
I found the problem, it was the ldap patch loading control files in the
wrong way in qmail-lspawn.
Franky
> ----------
> From: Harald Hanche-Olsen[SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 12:25 PM
> To: [EMAIL PROTECTED]
> Subject: Re: move the queue
>
> + Van Liedekerke Franky <[EMAIL PROTECTED]>:
>
> | 929610000.245483 alert: cannot start: hath the daemon spawn no fire?
>
> As far as I can tell from a quick look at the source, that means that
> either qmail-lspawn or qmail-rspawn fails to communicate. You might
> wish to trace system calls and see what is going on.
>
> | I followed the hints for moving the
> | queue like told in the README file provided with queue-fix.
> | Is it possible that qmail doesn't like a symbolic link?
>
> Not likely. qmail-send just changes to the queue directory, and
> hollers if it can't (a different kind of hollering than you saw).
>
> Actually, now that I think of it, I would suspect that something bad
> happened to the files in queue/lock when you moved the directory to a
> different file system. I am not sure if queue-fix fixes that?
>
> - Harald
>
<<< multipart/signed; boundary=T7mxYSe680VjQnyC; micalg=pgp-md5;protocol="application/pgp-signature": Unrecognized >>>
Thanks, not rebooting would cause the problem but I did reboot, I wish it
were that simple. I still get a disconnect when I execute telnet 127.0.0.1
25
My present theory is that I did not properly comment out the line in the
sendmail file in init.d
In the REMOVE.sendmail file the first step reads:
1. Find sendmail in your boot scripts. It's usually in either /etc/rc or
/etc/init.d/sendmail. It looks like
sendmail -bd -q15m
-q15m means that it should run the queue every 15 minutes; you may see a
different number. Comment out this line.
Well, my /etc/init.d/sendmail file does not contain the line, well, not
written out so nicely. It seems to build that line in a long set of
conditional statements building it up piece by piece. I thought I commented
out the right part but I'm not sure now.
The following is the contents of 2 files sendmail.~ (the original file) and
sendmail, my effort to comment out the appropriate line:
***************************************************************
sendmail.~ (the original file)
***************************************************************
# Check that networking is up.
[ ${NETWORKING} = "no" ] && exit 0
[ -f /usr/sbin/sendmail ] || exit 0
# See how we were called.
case "$1" in
start)
# Start daemons.
echo -n "Starting sendmail: "
daemon /usr/sbin/sendmail $([ "$DAEMON" = yes ] && echo -bd) \
$([ -n "$QUEUE" ] && echo -q$QUEUE)
echo
touch /var/lock/subsys/sendmail
;;
stop)
# Stop daemons.
echo -n "Shutting down sendmail: "
killproc sendmail
echo
rm -f /var/lock/subsys/sendmail
;;
restart)
$0 stop
$0 start
;;
status)
status sendmail
;;
*)
echo "Usage: sendmail {start|stop|restart|status}"
exit 1
esac
exit 0
***************************************************************
sendmail (my modified file)
***************************************************************
#!/bin/sh
#
# sendmail This shell script takes care of starting and stopping
# sendmail.
#
# chkconfig: 2345 80 30
# description: Sendmail is a Mail Transport Agent, which is the program \
# that moves mail from one machine to another.
# processname: sendmail
# config: /etc/sendmail.cf
# pidfile: /var/run/sendmail.pid
# Source function library.
. /etc/rc.d/init.d/functions
# Source networking configuration.
. /etc/sysconfig/network
# Source sendmail configureation.
if [ -f /etc/sysconfig/sendmail ] ; then
. /etc/sysconfig/sendmail
else
DAEMON=yes
QUEUE=1h
fi
# Check that networking is up.
[ ${NETWORKING} = "no" ] && exit 0
[ -f /usr/sbin/sendmail ] || exit 0
# See how we were called.
case "$1" in
start)
# Start daemons.
echo -n "Starting sendmail: "
# daemon /usr/sbin/sendmail $([
#"$DAEMON" = yes ] && echo -bd) \
# $([ -n
#"$QUEUE" ] && echo -q$QUEUE)
echo
touch /var/lock/subsys/sendmail
;;
stop)
# Stop daemons.
echo -n "Shutting down sendmail: "
killproc sendmail
echo
rm -f /var/lock/subsys/sendmail
;;
restart)
$0 stop
$0 start
;;
status)
status sendmail
;;
*)
echo "Usage: sendmail {start|stop|restart|status}"
exit 1
esac
exit 0
> -----Original Message-----
> From: Alex Miller [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 12:29 AM
> To: [EMAIL PROTECTED]
> Subject: TEST.receive failure
>
>
> I have run through the test.deliver tests and it seems okay, as near as I
> can tell.
>
> So, I did the remove of sendmail steps and set up qmail to start
> by itself.
> Everything seems ok, except for receiving remote mail
>
> I tried:
>
> TEST.recieve
>
> the first step is to run a telnet session.
>
> Here's what happens
> telnet 127.0.0.1 25
> Trying 127.0.0.1...
> Connected to 127.0.0.1
> Escapee character is '^]'
> Connection closed by foreign host.
>
> So somehow post 25 is failing.
>
> I tried sending myself an email remotely but it's been more than
> 5 hours and
> nothing has bounced back and I can't seem to see any indication in the
> /var/log/maillog of an attempt to deliver.
>
> Alex Miller
>
>
Correction: the file I was talking about is /etc/rc.d/init.d/sendmail not
/etc/initd/sendmail
> -----Original Message-----
> From: Alex Miller [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 8:54 AM
> To: [EMAIL PROTECTED]
> Subject: RE: TEST.receive failure
>
>
> Thanks, not rebooting would cause the problem but I did reboot, I
> wish it were that simple. I still get a disconnect when I execute
> telnet 127.0.0.1 25
> My present theory is that I did not properly comment out the line
> in the sendmail file in init.d
>
> In the REMOVE.sendmail file the first step reads:
>
> 1. Find sendmail in your boot scripts. It's usually in either
> /etc/rc or /etc/init.d/sendmail. It looks like
> sendmail -bd -q15m
> -q15m means that it should run the queue every 15 minutes; you
> may see a different number. Comment out this line.
>
> Well, my /etc/init.d/sendmail file does not contain the line,
> well, not written out so nicely. It seems to build that line in a
> long set of conditional statements building it up piece by piece.
> I thought I commented out the right part but I'm not sure now.
>
> The following is the contents of 2 files sendmail.~ (the original
> file) and sendmail, my effort to comment out the appropriate line:
>
> ***************************************************************
> sendmail.~ (the original file)
> ***************************************************************
> # Check that networking is up.
> [ ${NETWORKING} = "no" ] && exit 0
>
> [ -f /usr/sbin/sendmail ] || exit 0
>
> # See how we were called.
> case "$1" in
> start)
> # Start daemons.
> echo -n "Starting sendmail: "
> daemon /usr/sbin/sendmail $([ "$DAEMON" = yes ] && echo -bd) \
> $([ -n "$QUEUE" ] && echo -q$QUEUE)
> echo
> touch /var/lock/subsys/sendmail
> ;;
> stop)
> # Stop daemons.
> echo -n "Shutting down sendmail: "
> killproc sendmail
> echo
> rm -f /var/lock/subsys/sendmail
> ;;
> restart)
> $0 stop
> $0 start
> ;;
> status)
> status sendmail
> ;;
> *)
> echo "Usage: sendmail {start|stop|restart|status}"
> exit 1
> esac
>
> exit 0
>
> ***************************************************************
> sendmail (my modified file)
> ***************************************************************
> #!/bin/sh
> #
> # sendmail This shell script takes care of starting and stopping
> # sendmail.
> #
> # chkconfig: 2345 80 30
> # description: Sendmail is a Mail Transport Agent, which is the program \
> # that moves mail from one machine to another.
> # processname: sendmail
> # config: /etc/sendmail.cf
> # pidfile: /var/run/sendmail.pid
>
> # Source function library.
> . /etc/rc.d/init.d/functions
>
> # Source networking configuration.
> . /etc/sysconfig/network
>
> # Source sendmail configureation.
> if [ -f /etc/sysconfig/sendmail ] ; then
> . /etc/sysconfig/sendmail
> else
> DAEMON=yes
> QUEUE=1h
> fi
>
> # Check that networking is up.
> [ ${NETWORKING} = "no" ] && exit 0
>
> [ -f /usr/sbin/sendmail ] || exit 0
>
> # See how we were called.
> case "$1" in
> start)
> # Start daemons.
> echo -n "Starting sendmail: "
>
> # daemon /usr/sbin/sendmail $([
> #"$DAEMON" = yes ] && echo -bd) \
> # $([ -n
> #"$QUEUE" ] && echo -q$QUEUE)
> echo
> touch /var/lock/subsys/sendmail
> ;;
> stop)
> # Stop daemons.
> echo -n "Shutting down sendmail: "
> killproc sendmail
> echo
> rm -f /var/lock/subsys/sendmail
> ;;
> restart)
> $0 stop
> $0 start
> ;;
> status)
> status sendmail
> ;;
> *)
> echo "Usage: sendmail {start|stop|restart|status}"
> exit 1
> esac
>
> exit 0
>
>
> > -----Original Message-----
> > From: Alex Miller [mailto:[EMAIL PROTECTED]]
> > Sent: Thursday, June 17, 1999 12:29 AM
> > To: [EMAIL PROTECTED]
> > Subject: TEST.receive failure
> >
> >
> > I have run through the test.deliver tests and it seems okay, as
> near as I
> > can tell.
> >
> > So, I did the remove of sendmail steps and set up qmail to start
> > by itself.
> > Everything seems ok, except for receiving remote mail
> >
> > I tried:
> >
> > TEST.recieve
> >
> > the first step is to run a telnet session.
> >
> > Here's what happens
> > telnet 127.0.0.1 25
> > Trying 127.0.0.1...
> > Connected to 127.0.0.1
> > Escapee character is '^]'
> > Connection closed by foreign host.
> >
> > So somehow post 25 is failing.
> >
> > I tried sending myself an email remotely but it's been more than
> > 5 hours and
> > nothing has bounced back and I can't seem to see any indication in the
> > /var/log/maillog of an attempt to deliver.
> >
> > Alex Miller
> >
> >
Cody Payne wrote:
> Ok I have installed two instances of qMail each one bound to its specific
> NIC card w/ separate tcpserver instances one for qmail and one for qmail2.
> The issue I am having is that it seems that the /var/qmail2
> /alias/.qmail-x files in the second instance of qmail are not being read.
> But qMail2 is reading the .qmail files from the /var/qmail/alias/.qmail.
> Is this right? Can qMail support two instances of itself? How can I get
> e-mail coming in the second instance to read the correct .qmail file to
> send it to the proper local mailbox?
qmail doesn't look in $qmaildir/alias for alias .qmail files, it looks in
~alias. If you want your instances to have separate alias .qmail files, you'll
have to create 2 alias users and change conf-users to reflect this, then
recompile.
--
Peter Haworth [EMAIL PROTECTED]
"Can't Windows people think? Uhm, what did I just say? :-)"
-- Jarkko Hietaniemi
Sory for the inconvinience.
It looks like there is a bug in MS Outlook, where it is having trouble
displaying emails with certain header information. True enough, Microsoft is
evil incarnate but it does read 99.99% of email fairly well. I checked and
the Netscape client on my LINUX box reads the same emails just fine.
The emails Outlook is having trouble with have all been sent to this list.
All of the emails sent from Chris Garrigues and one email from Mark Drummond
(the one titled Re: qmail + Maildir + Procmail) have this problem.
My concern is that if there is something slightly dialectical about Chris's
and Mark's QMAIL setup, MS Outlook recipients will have trouble with those
emails.
Since I am setting up QMail as part of a not-for-profit ISP service that
provides internet services to participating not-for-profits, I would not
want my email system to send out email which is a dialect of normal email
which MS Outlook (and potentially other evil clients) can't handle.
I think it's worth looking into what is different about these emails and how
qmail is causing that difference.
Alex Miller
On Thu, Jun 17, 1999 at 10:57:50AM -0400, Alex Miller wrote:
} It looks like there is a bug in MS Outlook, where it is having trouble
} displaying emails with certain header information. True enough, Microsoft is
} evil incarnate but it does read 99.99% of email fairly well. I checked and
} the Netscape client on my LINUX box reads the same emails just fine.
}
} The emails Outlook is having trouble with have all been sent to this list.
} All of the emails sent from Chris Garrigues and one email from Mark Drummond
} (the one titled Re: qmail + Maildir + Procmail) have this problem.
}
} My concern is that if there is something slightly dialectical about Chris's
} and Mark's QMAIL setup, MS Outlook recipients will have trouble with those
} emails.
}
} Since I am setting up QMail as part of a not-for-profit ISP service that
} provides internet services to participating not-for-profits, I would not
} want my email system to send out email which is a dialect of normal email
} which MS Outlook (and potentially other evil clients) can't handle.
}
} I think it's worth looking into what is different about these emails and how
} qmail is causing that difference.
To hazard a guess, if I remember their messages in mutt this morning
and yesterday it's because both of them use PGP-MIME. Microsoft
simply doesn't do standards very well. You can't ask them not to use
PGP-MIME just because Microsoft crapware can't handle it.
}
} Alex Miller
}
--
--------
Paul J. Schinder
NASA Goddard Space Flight Center
[EMAIL PROTECTED]
>It looks like there is a bug in MS Outlook, where it is having trouble
>displaying emails with certain header information. True enough, Microsoft is
>evil incarnate but it does read 99.99% of email fairly well. I checked and
>the Netscape client on my LINUX box reads the same emails just fine.
>
>The emails Outlook is having trouble with have all been sent to this list.
>All of the emails sent from Chris Garrigues and one email from Mark Drummond
>(the one titled Re: qmail + Maildir + Procmail) have this problem.
If it's any help: the emails you referred to seem to have been
pgp-signed - I'm using a beta of Eudora 4.2 on a mac and have
installed the latest Freeware international PGP (6.02i), but, and I
suspect this is important, I've not configured PGP at all (no keyring
or anything). What I get is an attachment with a PGP icon -
double-clicking it gets me
"The PGP plugin could not find your keyring files. Please set up your
keyring files using PGPtools."
Hope that helps
John
Who's obviously going to have to set up his keyring...
--
John O'Shea <mailto:[EMAIL PROTECTED]>
Wordbank Ltd, 9 Orange St, London WC2H 7EA
T:+44 171 766 7200 F:+44 171 766 7222
Yes, the problem seems to be in the pane viewing. So, if using PGP-mime is a
standard that is well-recognized and MS Outlook doesn't handle it properly,
then it is an MS Outlook bug.
But I'm really surprised that I haven't encountered this problem before? Are
Chris and Mark the only folks out there using PGP? I subscribe to a lot of
mailing lists, which get shunted off to various folders, so I recieve mail
from literally thousands of users.
I haven't encountered the problem before so if it is strictly a MS Outlook
inability to understand PGP-mime properly I would expect to see the problem
in some significant percentage of emails I recieve, say 10%.
But this is a very small percentage, I find it hard to believe that
something as famous as PGP would be so underrepresented in all these mailing
lists.
Alex Miller
> -----Original Message-----
> From: Paul J. Schinder [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 11:05 AM
> To: Alex Miller
> Cc: List-QMAIL
> Subject: Re: Tower of Babel problem
>
>
> On Thu, Jun 17, 1999 at 10:57:50AM -0400, Alex Miller wrote:
> } It looks like there is a bug in MS Outlook, where it is having trouble
> } displaying emails with certain header information. True enough,
> Microsoft is
> } evil incarnate but it does read 99.99% of email fairly well. I
> checked and
> } the Netscape client on my LINUX box reads the same emails just fine.
> }
> } The emails Outlook is having trouble with have all been sent to
> this list.
> } All of the emails sent from Chris Garrigues and one email from
> Mark Drummond
> } (the one titled Re: qmail + Maildir + Procmail) have this problem.
> }
> } My concern is that if there is something slightly dialectical
> about Chris's
> } and Mark's QMAIL setup, MS Outlook recipients will have trouble
> with those
> } emails.
> }
> } Since I am setting up QMail as part of a not-for-profit ISP service that
> } provides internet services to participating not-for-profits, I would not
> } want my email system to send out email which is a dialect of
> normal email
> } which MS Outlook (and potentially other evil clients) can't handle.
> }
> } I think it's worth looking into what is different about these
> emails and how
> } qmail is causing that difference.
>
> To hazard a guess, if I remember their messages in mutt this morning
> and yesterday it's because both of them use PGP-MIME. Microsoft
> simply doesn't do standards very well. You can't ask them not to use
> PGP-MIME just because Microsoft crapware can't handle it.
>
> }
> } Alex Miller
> }
>
> --
> --------
> Paul J. Schinder
> NASA Goddard Space Flight Center
> [EMAIL PROTECTED]
>
On Thu, Jun 17, 1999 at 11:24:36AM -0400, Alex Miller wrote:
} Yes, the problem seems to be in the pane viewing. So, if using PGP-mime is a
} standard that is well-recognized and MS Outlook doesn't handle it properly,
} then it is an MS Outlook bug.
}
} But I'm really surprised that I haven't encountered this problem before? Are
} Chris and Mark the only folks out there using PGP? I subscribe to a lot of
} mailing lists, which get shunted off to various folders, so I recieve mail
} from literally thousands of users.
There are a couple of different ways that e-mail gets PGP signed.
PGP-MIME seems to be the more uncommon way, at least in the e-mail I
get. The usual way is to simply include the PGP signature in text at
the bottom of the message with some simple non-MIME delimiters around
the message (on this list Sam, if I remember right, signs his e-mail
in this way). Some software, like mutt, only does PGP-MIME signing.
Eudora can do it either way.
If you receive PGP-MIME mail using (Mac) Eudora, you get a little icon
instead of the message. You need to double click on the icon for
Eudora to check the signature and display the message (and then you
have to remember to drag the attachment to the Trash if you don't want
it lying around). The older format PGP signed mail simply displays,
and you need to go to extra trouble only if you want to verify the
signature.
}
} I haven't encountered the problem before so if it is strictly a MS Outlook
} inability to understand PGP-mime properly I would expect to see the problem
} in some significant percentage of emails I recieve, say 10%.
}
} But this is a very small percentage, I find it hard to believe that
} something as famous as PGP would be so underrepresented in all these mailing
} lists.
}
} Alex Miller
--
--------
Paul J. Schinder
NASA Goddard Space Flight Center
[EMAIL PROTECTED]
> On Thu, Jun 17, 1999 at 10:57:50AM -0400, Alex Miller wrote:
> } It looks like there is a bug in MS Outlook, where it is having trouble
> } displaying emails with certain header information.
I have no problem with Outlook Express. Is it just big-boy Outlook that
fails?
___________________________________________________________________
This message has been checked for all viruses, including ExploreZip
by the Star Screening System
http://academy.star.co.uk/public/virustats.htm
> From: Paul Schinder <[EMAIL PROTECTED]>
> Date: Thu, 17 Jun 1999 11:35:47 -0400
>
> On Thu, Jun 17, 1999 at 11:24:36AM -0400, Alex Miller wrote:
> } But I'm really surprised that I haven't encountered this problem before? Are
> } Chris and Mark the only folks out there using PGP? I subscribe to a lot of
> } mailing lists, which get shunted off to various folders, so I recieve mai
> } from literally thousands of users.
>
> There are a couple of different ways that e-mail gets PGP signed.
> PGP-MIME seems to be the more uncommon way, at least in the e-mail I
> get. The usual way is to simply include the PGP signature in text at
> the bottom of the message with some simple non-MIME delimiters around
> the message (on this list Sam, if I remember right, signs his e-mail
> in this way). Some software, like mutt, only does PGP-MIME signing.
> Eudora can do it either way.
Basically, the plain text approach isn't a standard, it's more of a
convention. It's the most compatible, but it doesn't allow you to sign
multipart or otherwise typed email messages. What usually happens if you have
an attachment is that the plain text is signed, but the attachment isn't. If
you think about it, that's kinda silly since the attachment is probably more
critical than the text which says something like "here's the file I promised
you."
The PGP-MIME approach is a standard (I don't recall which RFC off hand), but
it's not widely deployed yet. As a rule, most clients that don't support
PGP-MIME see the signature as an attachment, so I frequently get email from
people with older versions of Eudora asking me what that attachment was that I
sent them. (They usually tell me they tried to open it as well...I assume
these are the same people who pass email viruses around.)
Chris
--
Chris Garrigues virCIO
+1 512 432 4046 4314 Avenue C O-
http://www.DeepEddy.Com/~cwg/ Austin, TX 78751-3709
+1 512 374 0500
My email address is an experiment in SPAM elimination. For an
explanation of what we're doing, see http://www.DeepEddy.Com/tms.html
Nobody ever got fired for buying Microsoft,
but they could get fired for relying on Microsoft.
PGP signature
AFAIK "Internet Mail Service" means that the sender is using Microsoft
Exchange client to send the email.
I wouldn't recommend blocking all email coming from this client as you are
likely to block more legitimate mail than spam.
--Adam
On Thu, Jun 17, 1999 at 11:27:30AM +0200, torben fjerdingstad wrote:
> I my postmaster mailbox I get a lot of trash from those of
> our customers that use an MTA which identifies itself as:
> Internet Mail Service (5.5.2232.9) and
> Internet Mail Service (5.5.2448.0).
>
> That really a pain.
>
> How can I prevent those messages from ending up as double
> bounces in my postmaster mailbox? They are totally useless,
> even for spammer tracing, because the received lines are
> missing.
>
> Can I make qmail block them?
>
> I think the blocking criteria might be:
> If return-path is empty and from is [EMAIL PROTECTED], drop it.
>
> What would you do / what are you doing?
> Do you know the product so I can tell the customers how to fix
> their systems?
>
> Here is a typical message that ends up as a double bounce. In this
> case the bogus sender is a spammer. It could also originate from
> a non-spammer, and sometimes it is a positive delivery notification
> like "Your message was delivered to the following recipeint(s):":
>
> Return-Path: <>
> Received: (qmail 12368 invoked from network); 17 Jun 1999 07:10:22 -0000
> Received: from foulum01.agrsci.dk (130.226.240.22)
> by mail.net.uni-c.dk with SMTP; 17 Jun 1999 07:10:22 -0000
> Received: by foulum01.agrsci.dk with Internet Mail Service (5.5.2232.9)
> id <NBKR28MG>; Thu, 17 Jun 1999 09:09:14 +0200
> Message-ID: <[EMAIL PROTECTED]>
> From: System Administrator <[EMAIL PROTECTED]>
> To: [EMAIL PROTECTED]
> Subject: Undeliverable: "How to Pay Zero Taxes" Selling in stores for $30,Free!
> axwdm
> Date: Thu, 17 Jun 1999 09:09:13 +0200
> MIME-Version: 1.0
> X-Mailer: Internet Mail Service (5.5.2232.9)
> X-MS-Embedded-Report:
> Content-Type: multipart/mixed;
> boundary="----_=_NextPart_000_01BEB890.4E28FAE4"
>
> This message is in MIME format. Since your mail reader does not understand
> this format, some or all of this message may not be legible.
>
> ------_=_NextPart_000_01BEB890.4E28FAE4
> Content-Type: text/plain
>
> Your message
>
> To: [EMAIL PROTECTED]
> Subject: "How to Pay Zero Taxes" Selling in stores for $30,Free!
> axwdm
> Sent: Thu, 17 Jun 1999 09:06:55 +0200
>
> did not reach the following recipient(s):
>
> [EMAIL PROTECTED] on Thu, 17 Jun 1999 09:09:11 +0200
> The recipient name is not recognized
> MSEXCH:IMS:Danmarks JordbrugsForskning:Foulum:FOULUM01 0 (000C05A6)
> Unknown Recipient
>
> ------_=_NextPart_000_01BEB890.4E28FAE4
> Content-Type: message/rfc822
> Content-Location: ATT-0-223AA4ED5324D311997600805F8BB749
>
> Message-ID: <[EMAIL PROTECTED]>
> From: [EMAIL PROTECTED]
> To: [EMAIL PROTECTED]
> Subject: "How to Pay Zero Taxes" Selling in stores for $30,Free!
> axwdm
> Date: Thu, 17 Jun 1999 09:06:55 +0200
> MIME-Version: 1.0
> X-Mailer: Internet Mail Service (5.5.2232.9)
> X-MS-Embedded-Report:
> Content-Type: text/plain;
> charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
>
> [ body deleted by me ]
>
> ------_=_NextPart_000_01BEB890.4E28FAE4--
>
>
> --
> Med venlig hilsen / Regards
> Netdriftgruppen / Network Management Group
> UNI-C
>
> Tlf./Phone +45 35 87 89 41 Mail: UNI-C
> Fax. +45 35 87 89 90 Bygning 304
> E-mail: [EMAIL PROTECTED] DK-2800 Lyngby
>
On Thu, Jun 17, 1999 at 10:59:39AM -0400, Adam D. McKenna wrote:
> AFAIK "Internet Mail Service" means that the sender is using Microsoft
> Exchange client to send the email.
>
> I wouldn't recommend blocking all email coming from this client as you are
> likely to block more legitimate mail than spam.
Fred's solution is wonderful. qmail is wonderful.
I grep for "X-MS-Embedded-Report: ". That solution is just right :)
> On Thu, Jun 17, 1999 at 11:27:30AM +0200, torben fjerdingstad wrote:
> > I my postmaster mailbox I get a lot of trash from those of
> > our customers that use an MTA which identifies itself as:
> > Internet Mail Service (5.5.2232.9) and
> > Internet Mail Service (5.5.2448.0).
> > Can I make qmail block them?
> >
> > Return-Path: <>
> > Received: (qmail 12368 invoked from network); 17 Jun 1999 07:10:22 -0000
> > Received: from foulum01.agrsci.dk (130.226.240.22)
> > by mail.net.uni-c.dk with SMTP; 17 Jun 1999 07:10:22 -0000
> > Received: by foulum01.agrsci.dk with Internet Mail Service (5.5.2232.9)
> > id <NBKR28MG>; Thu, 17 Jun 1999 09:09:14 +0200
> > Message-ID: <[EMAIL PROTECTED]>
> > From: System Administrator <[EMAIL PROTECTED]>
> > To: [EMAIL PROTECTED]
--
Med venlig hilsen / Regards
Netdriftgruppen / Network Management Group
UNI-C
Tlf./Phone +45 35 87 89 41 Mail: UNI-C
Fax. +45 35 87 89 90 Bygning 304
E-mail: [EMAIL PROTECTED] DK-2800 Lyngby
On Wed, 16 Jun 1999, Albert Hopkins wrote:
AH>
AH>Using qmail, is /bin/mail supposed to be replaced (by a qmail-* program)?
/bin/mail and /bin/mailx and friends, if they expect to manipulate your
local copy of sendmail (be it in /usr/sbin, /usr/lib, whatever), should work
just fine, provided that you install qmail to use /var/spool/mail
(INSTALL.VSM) and symbolically link /var/qmail/bin/sendmail to your
sendmamil's original location. Or at least that's how we de-sendmailify
solaris boxen here.....
--
--Matt Schnierle
--mgs at stargate dot net
--Stargate Industries, LLC
--#include <std/disclaimer.h>
--"It's not that simple."
When telnetting to port 25, why do I have to type an extra ENTER to
disconnect from qmail-smtpd? Is this an RFC-compliance issue?
telnet mail.fwi.com 25
Trying 209.84.160.144...
Connected to mail.fwi.com.
Escape character is '^]'.
220 mail.fwi.com ESMTP Sendmail 8.8.5/8.8.5; Thu, 17 Jun 1999
11:07:22 -0500 (ES
T)
quit <== I type
221 mail.fwi.com closing connection
Connection closed by foreign host.
telnet faxsrv.fwi.com 25
Trying 209.84.160.13...
Connected to faxsrv.fwi.com.
Escape character is '^]'.
220 FWI mail server - qmail 1.03 ESMTP
quit <== I type
221 FWI mail server - qmail 1.03
<== I also type ENTER (LF?) here (not necessary for
sendmail)
Connection closed by foreign host.
- - - - - - - - - - - - - - - - - - - - - - -
Peeter Pirn, Sys Admin, FWI Internet, 219-426-7701 x17
On Thu, 17 Jun 1999, it was written:
> When telnetting to port 25, why do I have to type an extra ENTER to
> disconnect from qmail-smtpd? Is this an RFC-compliance issue?
> [...]
> telnet faxsrv.fwi.com 25
> Trying 209.84.160.13...
> Connected to faxsrv.fwi.com.
> Escape character is '^]'.
> 220 FWI mail server - qmail 1.03 ESMTP
> quit <== I type
> 221 FWI mail server - qmail 1.03
> <== I also type ENTER (LF?) here (not necessary for
> sendmail)
> Connection closed by foreign host.
My host does
krondor.kva.hu:[pts/1]~>>% telnet krondor.kva.hu 25
Trying 195.228.124.65...
Connected to krondor.kva.hu.
Escape character is '^]'.
220 krondor.kva.hu ESMTP
quit
221 krondor.kva.hu
Connection closed by foreign host.
It works without any problems. What OS do you run?
--
Regards: Kevin (Balazs)
this happens for me too, using the fixcr method of fixing bare LFs from the
qmail page.
On Thu, Jun 17, 1999 at 07:38:50PM +0200, Balazs Nagy wrote:
# On Thu, 17 Jun 1999, it was written:
#
# > When telnetting to port 25, why do I have to type an extra ENTER to
# > disconnect from qmail-smtpd? Is this an RFC-compliance issue?
# > [...]
# > telnet faxsrv.fwi.com 25
# > Trying 209.84.160.13...
# > Connected to faxsrv.fwi.com.
# > Escape character is '^]'.
# > 220 FWI mail server - qmail 1.03 ESMTP
# > quit <== I type
# > 221 FWI mail server - qmail 1.03
# > <== I also type ENTER (LF?) here (not necessary for
# > sendmail)
# > Connection closed by foreign host.
#
# My host does
#
# krondor.kva.hu:[pts/1]~>>% telnet krondor.kva.hu 25
# Trying 195.228.124.65...
# Connected to krondor.kva.hu.
# Escape character is '^]'.
# 220 krondor.kva.hu ESMTP
# quit
# 221 krondor.kva.hu
# Connection closed by foreign host.
#
# It works without any problems. What OS do you run?
# --
# Regards: Kevin (Balazs)
#
--
/- [EMAIL PROTECTED] --------------- [EMAIL PROTECTED] -\
|Justin Bell NIC:JB3084| Time and rules are changing. |
|Pearson | Attention span is quickening. |
|Developer | Welcome to the Information Age. |
\-------- http://www.superlibrary.com/people/justin/ ----------/
| When telnetting to port 25, why do I have to type an extra ENTER to
| disconnect from qmail-smtpd? Is this an RFC-compliance issue?
No. Something's weird about your machine. It doesn't seem to happen
with any other qmail host that I tried. It's not your telnet client
either - I checked the machine you mentioned, and observed the same
phenomenon.
- Harald
On Thu, 17 Jun 1999, Justin Bell wrote:
> this happens for me too, using the fixcr method of fixing bare LFs from the
> qmail page.
Yes. It's a problem with fixcr. Just try it out as running 'qmail-smtpd'
then 'fixcr | qmail-smtpd' and fed them with a simple quit command.
Fixcr is bogus.
--
Regards: Kevin (Balazs)
You could telnet to my port 25 and you would be disconnected immediately, no
extra carriage return required ;-)
I'm struggling with getting my system running with QMail for this very
reason.
> -----Original Message-----
> From: Peeter Pirn [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 12:21 PM
> To: [EMAIL PROTECTED]
> Subject: Closing SMTP connection
>
>
> When telnetting to port 25, why do I have to type an extra ENTER to
> disconnect from qmail-smtpd? Is this an RFC-compliance issue?
>
> telnet mail.fwi.com 25
> Trying 209.84.160.144...
> Connected to mail.fwi.com.
> Escape character is '^]'.
> 220 mail.fwi.com ESMTP Sendmail 8.8.5/8.8.5; Thu, 17 Jun 1999
> 11:07:22 -0500 (ES
> T)
> quit <== I type
> 221 mail.fwi.com closing connection
> Connection closed by foreign host.
>
> telnet faxsrv.fwi.com 25
> Trying 209.84.160.13...
> Connected to faxsrv.fwi.com.
> Escape character is '^]'.
> 220 FWI mail server - qmail 1.03 ESMTP
> quit <== I type
> 221 FWI mail server - qmail 1.03
> <== I also type ENTER (LF?) here (not necessary for
> sendmail)
> Connection closed by foreign host.
>
> - - - - - - - - - - - - - - - - - - - - - - -
> Peeter Pirn, Sys Admin, FWI Internet, 219-426-7701 x17
>
>
>
Balazs Nagy wrote/schrieb/scribsit:
> Yes. It's a problem with fixcr. Just try it out as running 'qmail-smtpd'
> then 'fixcr | qmail-smtpd' and fed them with a simple quit command.
>
> Fixcr is bogus.
No, but doing s/qmail-smtpd/fixcr | qmail-smtpd/ is.
Run 'sh -c "fixcr | qmail-smtpd"' instead.
Harald H.-O.: wouldn't 'pipe' do well here, too?
Stefan
On Thu, Jun 17, 1999 at 07:13:21PM +0200, Stefan Paletta wrote:
# Balazs Nagy wrote/schrieb/scribsit:
# > Yes. It's a problem with fixcr. Just try it out as running 'qmail-smtpd'
# > then 'fixcr | qmail-smtpd' and fed them with a simple quit command.
# >
# > Fixcr is bogus.
#
# No, but doing s/qmail-smtpd/fixcr | qmail-smtpd/ is.
# Run 'sh -c "fixcr | qmail-smtpd"' instead.
#
so,
/usr/local/bin/tcpserver -v -x /etc/smtpd.cdb 0 80025 sh -c '/usr/local/bin/fixcr |
/var/qmail/bin/qmail-smtpd' 2>&1 | /usr/local/bin/accustamp| /usr/local/bin/cyclog
-s104000 -n5 /var/qmail/cf-smtpd-log/ &
should work perfectly then?
try telnetting to iq-ss5.iquest.net 80025
you will have to quit and then hit enter
--
/- [EMAIL PROTECTED] --------------- [EMAIL PROTECTED] -\
|Justin Bell NIC:JB3084| Time and rules are changing. |
|Pearson | Attention span is quickening. |
|Developer | Welcome to the Information Age. |
\-------- http://www.superlibrary.com/people/justin/ ----------/
Stefan Paletta <[EMAIL PROTECTED]> writes:
| Run 'sh -c "fixcr | qmail-smtpd"' instead.
No, because just piping to smtpd is insufficient for fixcr to be
informed when smtp exits. That's why you need the second, anomalous,
line of input: so that fixcr can write to a broken pipe and die.
Thanks! Yep. Fixcr caused it for me, too (BSD/OS 4.0):
su-2.01# rcsdiff -c -r1.1 -r1.2 rc.local
===================================================================
RCS file: RCS/rc.local,v
retrieving revision 1.1
retrieving revision 1.2
diff -c -r1.1 -r1.2
*** rc.local 1999/06/09 15:45:47 1.1
--- rc.local 1999/06/17 17:00:01 1.2
***************
*** 65,69 ****
csh -cf '/var/qmail/rc &'
# Start tcpserver running qmail-smtpd.
tcpserver -x /etc/tcp.smtp.cdb -v -u 7770 -g 2108 0 smtp \
! /var/qmail/bin/qmail-smtpd 2>&1 | /var/qmail/bin/splogger \
! smtpd 3 &
--- 65,69 ----
csh -cf '/var/qmail/rc &'
# Start tcpserver running qmail-smtpd.
tcpserver -x /etc/tcp.smtp.cdb -v -u 7770 -g 2108 0 smtp \
! sh -c '/usr/local/bin/fixcr | /var/qmail/bin/qmail-smtpd' 2>&1
\
! | /var/qmail/bin/splogger smtpd 3 &
Question: In the telnet interaction, why didn't the connection close?
I presume that the LF I sent after QUIT was received by tcpserver, fed
to fixcr, converted to CRLF and piped to qmail-smtpd. This should be
fine for QUIT like it is for any other SMTP command (They work okay.).
The next question is: Does using fixcr cause timeout problems for
other mail clients? I ran mail -v from the command line and there was
no delay in closing the connection. If I find any problems, I'll post.
> -----Original Message-----
> From: Justin Bell [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 11:42 AM
> To: Balazs Nagy
> Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]
> Subject: Re: Closing SMTP connection
>
>
> this happens for me too, using the fixcr method of fixing
> bare LFs from the
> qmail page.
>
>
> On Thu, Jun 17, 1999 at 07:38:50PM +0200, Balazs Nagy wrote:
> # On Thu, 17 Jun 1999, it was written:
> #
> # > When telnetting to port 25, why do I have to type an
> extra ENTER to
> # > disconnect from qmail-smtpd? Is this an RFC-compliance issue?
> # > [...]
> # > telnet faxsrv.fwi.com 25
> # > Trying 209.84.160.13...
> # > Connected to faxsrv.fwi.com.
> # > Escape character is '^]'.
> # > 220 FWI mail server - qmail 1.03 ESMTP
> # > quit <== I type
> # > 221 FWI mail server - qmail 1.03
> # > <== I also type ENTER (LF?) here (not necessary
for
> # > sendmail)
> # > Connection closed by foreign host.
> #
> # My host does
> #
> # krondor.kva.hu:[pts/1]~>>% telnet krondor.kva.hu 25
> # Trying 195.228.124.65...
> # Connected to krondor.kva.hu.
> # Escape character is '^]'.
> # 220 krondor.kva.hu ESMTP
> # quit
> # 221 krondor.kva.hu
> # Connection closed by foreign host.
> #
> # It works without any problems. What OS do you run?
> # --
> # Regards: Kevin (Balazs)
> #
>
> --
> /- [EMAIL PROTECTED] --------------- [EMAIL PROTECTED] -\
> |Justin Bell NIC:JB3084| Time and rules are changing. |
> |Pearson | Attention span is quickening. |
> |Developer | Welcome to the Information Age. |
> \-------- http://www.superlibrary.com/people/justin/ ----------/
>
unfortunately Cold Fusion barfs on it....
On Thu, Jun 17, 1999 at 12:53:01PM -0500,
[EMAIL PROTECTED] wrote:
# Thanks! Yep. Fixcr caused it for me, too (BSD/OS 4.0):
#
# su-2.01# rcsdiff -c -r1.1 -r1.2 rc.local
# ===================================================================
# RCS file: RCS/rc.local,v
# retrieving revision 1.1
# retrieving revision 1.2
# diff -c -r1.1 -r1.2
# *** rc.local 1999/06/09 15:45:47 1.1
# --- rc.local 1999/06/17 17:00:01 1.2
# ***************
# *** 65,69 ****
# csh -cf '/var/qmail/rc &'
# # Start tcpserver running qmail-smtpd.
# tcpserver -x /etc/tcp.smtp.cdb -v -u 7770 -g 2108 0 smtp \
# ! /var/qmail/bin/qmail-smtpd 2>&1 | /var/qmail/bin/splogger \
# ! smtpd 3 &
# --- 65,69 ----
# csh -cf '/var/qmail/rc &'
# # Start tcpserver running qmail-smtpd.
# tcpserver -x /etc/tcp.smtp.cdb -v -u 7770 -g 2108 0 smtp \
# ! sh -c '/usr/local/bin/fixcr | /var/qmail/bin/qmail-smtpd' 2>&1
# \
# ! | /var/qmail/bin/splogger smtpd 3 &
#
# Question: In the telnet interaction, why didn't the connection close?
# I presume that the LF I sent after QUIT was received by tcpserver, fed
# to fixcr, converted to CRLF and piped to qmail-smtpd. This should be
# fine for QUIT like it is for any other SMTP command (They work okay.).
#
# The next question is: Does using fixcr cause timeout problems for
# other mail clients? I ran mail -v from the command line and there was
# no delay in closing the connection. If I find any problems, I'll post.
#
# > -----Original Message-----
# > From: Justin Bell [mailto:[EMAIL PROTECTED]]
# > Sent: Thursday, June 17, 1999 11:42 AM
# > To: Balazs Nagy
# > Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]
# > Subject: Re: Closing SMTP connection
# >
# >
# > this happens for me too, using the fixcr method of fixing
# > bare LFs from the
# > qmail page.
# >
# >
# > On Thu, Jun 17, 1999 at 07:38:50PM +0200, Balazs Nagy wrote:
# > # On Thu, 17 Jun 1999, it was written:
# > #
# > # > When telnetting to port 25, why do I have to type an
# > extra ENTER to
# > # > disconnect from qmail-smtpd? Is this an RFC-compliance issue?
# > # > [...]
# > # > telnet faxsrv.fwi.com 25
# > # > Trying 209.84.160.13...
# > # > Connected to faxsrv.fwi.com.
# > # > Escape character is '^]'.
# > # > 220 FWI mail server - qmail 1.03 ESMTP
# > # > quit <== I type
# > # > 221 FWI mail server - qmail 1.03
# > # > <== I also type ENTER (LF?) here (not necessary
# for
# > # > sendmail)
# > # > Connection closed by foreign host.
# > #
# > # My host does
# > #
# > # krondor.kva.hu:[pts/1]~>>% telnet krondor.kva.hu 25
# > # Trying 195.228.124.65...
# > # Connected to krondor.kva.hu.
# > # Escape character is '^]'.
# > # 220 krondor.kva.hu ESMTP
# > # quit
# > # 221 krondor.kva.hu
# > # Connection closed by foreign host.
# > #
# > # It works without any problems. What OS do you run?
# > # --
# > # Regards: Kevin (Balazs)
# > #
# >
# > --
# > /- [EMAIL PROTECTED] --------------- [EMAIL PROTECTED] -\
# > |Justin Bell NIC:JB3084| Time and rules are changing. |
# > |Pearson | Attention span is quickening. |
# > |Developer | Welcome to the Information Age. |
# > \-------- http://www.superlibrary.com/people/justin/ ----------/
# >
#
--
/- [EMAIL PROTECTED] --------------- [EMAIL PROTECTED] -\
|Justin Bell NIC:JB3084| Time and rules are changing. |
|Pearson | Attention span is quickening. |
|Developer | Welcome to the Information Age. |
\-------- http://www.superlibrary.com/people/justin/ ----------/
+ Stefan Paletta <[EMAIL PROTECTED]>:
| Balazs Nagy wrote/schrieb/scribsit:
| > Yes. It's a problem with fixcr. Just try it out as running 'qmail-smtpd'
| > then 'fixcr | qmail-smtpd' and fed them with a simple quit command.
| >
| > Fixcr is bogus.
|
| No, but doing s/qmail-smtpd/fixcr | qmail-smtpd/ is.
| Run 'sh -c "fixcr | qmail-smtpd"' instead.
|
| Harald H.-O.: wouldn't 'pipe' do well here, too?
Having tcpserver control pipe fixcr '|' qmail-smtpd rather than
sh -c 'fixcr | qmail-smtpd' does indeed save a fork and an extra
process per incoming connection.
But as far as I can see, none of these help against the problem that
started this thread. That problem appears to be the following:
fixcr copies stdin to stdout (fixing isolated newslines) until it
reaches end-of-file on stdin, or until it's killed by SIGPIPE.
When the client sends "quit\r\n", that is passed on to qmail-smtp,
which promptly prints its exit message and quits. However, fixcr will
not get its SIGPIPE until it actually tries to write more data to the
pipe. Hence the need for an extra return.
The best fix is to replace fixcr by a program that sets up the
pipeline and runs a command itself, to be used as follows:
newfixcr qmail-smtpd
And then it should expext SIGCHLD and die upon receiving it.
Adding the necessary code to fixcr.c in order to do this ought to be
rather trivial.
- Harald
+ Harald Hanche-Olsen <[EMAIL PROTECTED]>:
| Adding the necessary code to fixcr.c in order to do this ought to be
| rather trivial.
In fact, I include below a first attempt at such a fix.
I warn you that I have only superficially tested this!
This is not DJB-ware by any stretch of the imagination, but it just
might work - a preliminary test indicates that it does.
Apply this patch to ucspitcp 0.84, run make fixcr, and voila.
- Harald
diff -u /stores/stash/ucspitcp/src-0.84/Makefile ./Makefile
--- /stores/stash/ucspitcp/src-0.84/Makefile Thu Nov 12 06:32:01 1998
+++ ./Makefile Thu Jun 17 20:19:08 1999
@@ -252,8 +252,8 @@
nroff -man [email protected] > [email protected]
fixcr: \
-load fixcr.o substdio.a error.a str.a
- ./load fixcr substdio.a error.a str.a
+load fixcr.o substdio.a error.a str.a sig.a fd.a
+ ./load fixcr substdio.a error.a str.a sig.a fd.a
fixcr.0: \
fixcr.1
diff -u /stores/stash/ucspitcp/src-0.84/fixcr.1 ./fixcr.1
--- /stores/stash/ucspitcp/src-0.84/fixcr.1 Thu Nov 12 06:32:01 1998
+++ ./fixcr.1 Thu Jun 17 20:28:24 1999
@@ -3,9 +3,24 @@
fixcr \- make sure that there is a CR before each LF
.SH SYNOPSIS
.B fixcr
+[
+.I program
+[
+.I arg ...
+]]
.SH DESCRIPTION
.B fixcr
inserts CR at the end of each line of input where a CR is not already present.
It does not insert CR at the end of a partial final line.
+
+If
+.I program
+is specified,
+.B fixcr
+runs it with the given arguments and pipes its output into it.
+
+Otherwise,
+.B fixcr
+writes to standard output.
.SH "SEE ALSO"
addcr(1)
diff -u /stores/stash/ucspitcp/src-0.84/fixcr.c ./fixcr.c
--- /stores/stash/ucspitcp/src-0.84/fixcr.c Thu Nov 12 06:32:01 1998
+++ ./fixcr.c Thu Jun 17 21:12:27 1999
@@ -1,13 +1,37 @@
+#include "sig.h"
#include "substdio.h"
#include "subfd.h"
#include "exit.h"
-void main()
+static void die() { _exit(0); }
+
+void main(argc, argv)
+ int argc;
+ char *argv[];
{
register int n;
register char *x;
char ch;
char lastch = 0;
+ int p[2];
+
+ if (argc > 1)
+ {
+ if (pipe(&p)<0) _exit(111);
+ switch(fork())
+ {
+ case -1: _exit(111);
+ case 0:
+ if (fd_move(0,p[0])<0) die();
+ close(p[1]);
+ execvp(argv[1], argv + 1);
+ _exit(111);
+ default:
+ if (fd_move(1,p[1])<0) die();
+ close(p[0]);
+ sig_childunblock(); sig_childcatch(&die);
+ }
+ }
for (;;) {
n = substdio_feed(subfdin);
Hello List,
We're having a problem with a mailserver. The machine does about 11,000
msgs a day. At heaviest times smtp concurrency goes up to 60/150 or
70/150. During these times load avg goes up to 20, and syslog is eating
%70 of CPU. At that point the messages and maillog files are at say 20M
each. If we dev/null the logs, load returns to 1 2 or 3 load avg.
I saw a few articles in the archives about this. Seems that syslog isn't
up to the task. Comments, on whether syslog is the culprit, and what
logging mechanism to switch to welcome?
Server is a PII 128MB 12GB RH5.0 running tcpserver
Many Thanks - eric
+ + + + + + + + + + + + + + + + + + + +
Spark Sistemas E-mail
- presentado por IWCC Argentina S.A.
Tel: 4702-1958
e-mail: [EMAIL PROTECTED]
+ + + + + + + + + + + + + + + + + + + +
On Thu, 17 Jun 1999, Eric Dahnke wrote:
> We're having a problem with a mailserver. The machine does about 11,000
> msgs a day. At heaviest times smtp concurrency goes up to 60/150 or
> 70/150. During these times load avg goes up to 20, and syslog is eating
> %70 of CPU. At that point the messages and maillog files are at say 20M
> each. If we dev/null the logs, load returns to 1 2 or 3 load avg.
If you use tcpserver I suggest you to use cyclog instead. By the way your
problem is syslog's writing method. Linux's syslog does a sync after each
line it writes. If you put a dash (-) before the mail log file (eg.
/var/log/maillog), it won't sync. This isn't a perfect solution but it
works if your system doesn't crash every week.
--
Regards: Kevin (Balazs)
+ Eric Dahnke <[EMAIL PROTECTED]>:
| Comments, on whether syslog is the culprit,
Sure sounds like it from your description.
| and what logging mechanism to switch to welcome?
There is always cyclog, from Dan's daemontools package.
I use it, and am quite happy with it.
- Harald
On Thu, 17 Jun 1999, Eric Dahnke wrote:
> I saw a few articles in the archives about this. Seems that syslog isn't
> up to the task. Comments, on whether syslog is the culprit, and what
> logging mechanism to switch to welcome?
>
> Server is a PII 128MB 12GB RH5.0 running tcpserver
yes, syslog is probably the cuprit. fixes:
- use cyclog instead as others have suggested.
- put a '-' in front of the pathname in the syslogd.cong file in /etc to
prevent it from syncing to disk everytime qmail-* sends it a message
(if your installed syslog supports it -- my RH4.1++ system does.)
RjL
[++]with much of the original distribution replaced...
==================================================================
The problems of the world || Fax: +44 870 0521198
can't be solved by fixing || Email: [EMAIL PROTECTED]
the working -- C. Daniluk || Phone: +44 385 275 394
On Thu, 17 Jun 1999, Eric Dahnke wrote:
ED>Hello List,
ED>
ED>We're having a problem with a mailserver. The machine does about 11,000
ED>msgs a day. At heaviest times smtp concurrency goes up to 60/150 or
ED>70/150. During these times load avg goes up to 20, and syslog is eating
ED>%70 of CPU. At that point the messages and maillog files are at say 20M
ED>each. If we dev/null the logs, load returns to 1 2 or 3 load avg.
ED>
ED>I saw a few articles in the archives about this. Seems that syslog isn't
ED>up to the task. Comments, on whether syslog is the culprit, and what
ED>logging mechanism to switch to welcome?
Check out cyclog-part of the daemontools package:
ftp://koobera.math.uic.edu/www/daemontools.html
--
--Matt Schnierle
--mgs at stargate dot net
--Stargate Industries, LLC
--#include <std/disclaimer.h>
--"It's not that simple."
Now that I have qmail setup I am trying to work through some relaying problems
(most likely operator headspace error on my part).
Here's the scenario: I have clients on multiple class C's that need to send
and receive mail through the mail server. I've found that when mail is being
sent from a client, I need to have an entry in /etc/hosts.allow like
tcp-env: ALL: setenv = RELAYCLIENT
to let mail be sent to hosts that are not in the rcpthosts file. Is this
correct?
What concerns me is that it would seem that anyone can relay through the
server when it is setup that way. When I try using /etc/hosts.allow of the
form:
tcp-env: 1.2.3.: setenv = RELAYCLIENT
then the system will not allow any connections from the outside to the smtp
daemon so no incoming mail is allowed.
I must be missing something obvious and could use some assistance if anyone
can take a moment to point me in the correct direction.
Thanks.
-jeff
Jeffrey Finkelstein wrote/schrieb/scribsit:
> What concerns me is that it would seem that anyone can relay through the
> server when it is setup that way. When I try using /etc/hosts.allow of the
> form:
>
> tcp-env: 1.2.3.: setenv = RELAYCLIENT
>
> then the system will not allow any connections from the outside to the smtp
> daemon so no incoming mail is allowed.
You probably need to allow connections from everywhere then.
Either add
tcp-env: ALL
to hosts.allow or check if you have an
ALL: ALL
in hosts.deny and need it.
Could also be that tcpd denies by default, who knows with tcpd...
Stefan
Putting the ``tcp-env: ALL'' in the /etc/hosts.allow solves the incoming mail
problem, but clients cannot send mail through the mail host since the
destination mail address is not in the rcpthosts file.
I must be missing something obvious somewhere, but I'll be darned if I see it
just yet.
-jeff
On Thu, Jun 17, 1999 at 07:55:28PM +0200, Stefan Paletta wrote:
> Jeffrey Finkelstein wrote/schrieb/scribsit:
> > What concerns me is that it would seem that anyone can relay through the
> > server when it is setup that way. When I try using /etc/hosts.allow of the
> > form:
> >
> > tcp-env: 1.2.3.: setenv = RELAYCLIENT
> >
> > then the system will not allow any connections from the outside to the smtp
> > daemon so no incoming mail is allowed.
>
> You probably need to allow connections from everywhere then.
> Either add
>
> tcp-env: ALL
>
> to hosts.allow or check if you have an
>
> ALL: ALL
>
> in hosts.deny and need it.
>
> Could also be that tcpd denies by default, who knows with tcpd...
>
> Stefan
Jeffrey Finkelstein wrote/schrieb/scribsit:
> Putting the ``tcp-env: ALL'' in the /etc/hosts.allow solves the incoming mail
> problem, but clients cannot send mail through the mail host since the
> destination mail address is not in the rcpthosts file.
Your'e supposed to have both lines in hosts.allow. Like:
tcp-env: 10.0.0.0/255.0.0.0: setenv = RELAYCLIENT
tcp-env: ALL
I just tried it... but then again there are so many versions
of tcpd...
Stefan
You need BOTH lines in /etc/hosts.allow!!!!!
tcp-env: 1.2.3.: setenv = RELAYCLIENT
tcp-env: ALL
On Thu, 17 Jun 1999, Jeffrey Finkelstein wrote:
> Putting the ``tcp-env: ALL'' in the /etc/hosts.allow solves the incoming mail
> problem, but clients cannot send mail through the mail host since the
> destination mail address is not in the rcpthosts file.
>
> I must be missing something obvious somewhere, but I'll be darned if I see it
> just yet.
>
> -jeff
>
> On Thu, Jun 17, 1999 at 07:55:28PM +0200, Stefan Paletta wrote:
> > Jeffrey Finkelstein wrote/schrieb/scribsit:
> > > What concerns me is that it would seem that anyone can relay through the
> > > server when it is setup that way. When I try using /etc/hosts.allow of the
> > > form:
> > >
> > > tcp-env: 1.2.3.: setenv = RELAYCLIENT
> > >
> > > then the system will not allow any connections from the outside to the smtp
> > > daemon so no incoming mail is allowed.
> >
> > You probably need to allow connections from everywhere then.
> > Either add
> >
> > tcp-env: ALL
> >
> > to hosts.allow or check if you have an
> >
> > ALL: ALL
> >
> > in hosts.deny and need it.
> >
> > Could also be that tcpd denies by default, who knows with tcpd...
> >
> > Stefan
>
---------------------------------
Timothy L. Mayo mailto:[EMAIL PROTECTED]
Senior Systems Administrator
localconnect(sm)
http://www.localconnect.net/
The National Business Network Inc. http://www.nb.net/
One Monroeville Center, Suite 850
Monroeville, PA 15146
(412) 810-8888 Phone
(412) 810-8886 Fax
Hello, I need a bit of help. There is a mail that periodically come in a Mailbox
There is a loop that I need to break, but I don't know how.
Follow the cut & paste of the header, any suggestions ?
Alberto
---------------------------
Return-Path: <>
Delivered-To: [EMAIL PROTECTED]
Received: (qmail 15308 invoked from network); 17 Jun 1999 17:28:34 -0000
Received: from unknown (HELO mail.numerica.it) (212.210.176.130)
by ns.numerica.it with SMTP; 17 Jun 1999 17:28:34 -0000
Received: (qmail 10052 invoked by uid 1001); 17 Jun 1999 17:11:40 -0000
Delivered-to: [EMAIL PROTECTED]
Delivered-through: [EMAIL PROTECTED]
Received: (qmail 10046 invoked from network); 17 Jun 1999 17:11:36 -0000
Received: from unknown (HELO server1.condorgroup.it) (212.38.39.161)
by mail.numerica.it with SMTP; 17 Jun 1999 17:11:36 -0000
Received: from server1.condorgroup.com (server1.condorgroup.com [128.10.10.2]) by
server1.condorgroup.it (NTMail
4.01.0014/NY1604.00.d974751a) with ESMTP id akibaaaa for <[EMAIL PROTECTED]>; Thu,
17 Jun 1999 19:35:03 +0100
From: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
Subject: Failed mail: non-delivery
Date: Thu, 17 Jun 1999 19:35:03 +0100
Message-Id: <[EMAIL PROTECTED]>
X-Mailer: NTMail v4.01.0014
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="[EMAIL PROTECTED]==_"
This is a MIME-encapsulated message
[EMAIL PROTECTED]==_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
This mail message has not been delivered.
[EMAIL PROTECTED]==_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
The SMTP protocol error was:
500 Bouncing mail.
[EMAIL PROTECTED]==_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
The requested destination was:
[EMAIL PROTECTED]
[EMAIL PROTECTED]==_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
The message headers and first 50 body lines follow:
[EMAIL PROTECTED]==_
Content-Type: message/rfc822
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Return-Path: <[EMAIL PROTECTED]>
Delivered-To: [EMAIL PROTECTED]
Received: (qmail 3105 invoked from network); 25 May 1999 12:15:52 -0000
Received: from mail.numerica.it (212.210.176.130)
by ns.numerica.it with SMTP; 25 May 1999 12:15:52 -0000
Received: (qmail 27217 invoked by uid 1001); 25 May 1999 11:59:48 -0000
Delivered-to: [EMAIL PROTECTED]
Delivered-through: [EMAIL PROTECTED]
Received: (qmail 27211 invoked from network); 25 May 1999 11:59:47 -0000
Received: from ns.numerica.it (194.184.189.10)
by mail.numerica.it with SMTP; 25 May 1999 11:59:47 -0000
Received: (qmail 3013 invoked from network); 25 May 1999 12:15:28 -0000
Received: from rovato-a28.numerica.it (HELO localhost) (195.120.249.124)
by ns.numerica.it with SMTP; 25 May 1999 12:15:28 -0000
Received: (qmail 2625 invoked by uid 511); 25 May 1999 11:11:58 -0000
Received: (qmail 2616 invoked from network); 25 May 1999 11:11:55 -0000
Received: from localhost ([EMAIL PROTECTED])
by localhost with SMTP; 25 May 1999 11:11:55 -0000
Delivered-To: [EMAIL PROTECTED]
Received: from ns.numerica.it
by localhost with POP3 (fetchmail-4.7.1)
for mailin@localhost (single-drop); Tue, 25 May 1999 11:11:55 +0000 (UTC)
Received: (qmail 21037 invoked from network); 25 May 1999 11:09:50 -0000
Received: from mail.numerica.it (212.210.176.130)
by ns.numerica.it with SMTP; 25 May 1999 11:09:50 -0000
Received: (qmail 24448 invoked by uid 1001); 25 May 1999 10:53:48 -0000
Delivered-to: [EMAIL PROTECTED]
Delivered-through: [EMAIL PROTECTED]
Received: (qmail 24371 invoked from network); 25 May 1999 10:53:38 -0000
Received: from ns.numerica.it (194.184.189.10)
by mail.numerica.it with SMTP; 25 May 1999 10:53:38 -0000
Received: (qmail 16514 invoked from network); 25 May 1999 10:42:46 -0000
Received: from brescia-b17.numerica.it (HELO marco) (212.38.35.150)
by ns.numerica.it with SMTP; 25 May 1999 10:42:46 -0000
Message-ID: <000e01bea69b$5c770c00$962326d4@marco>
From: "TeknoSistem s.r.l." <[EMAIL PROTECTED]>
To: "\"Zampedri Auto\"" <[EMAIL PROTECTED]>, "\"yokohama\"" <[EMAIL PROTECTED]>,
"\"Xplab\"" <[EMAIL PROTECTED]>, "\"World Link\"" <[EMAIL PROTECTED]>,
"\"Vigilianza Notturna BS\"" <[EMAIL PROTECTED]>,
"\"Vigasio\"" <[EMAIL PROTECTED]>, "\"Verincolor\"" <[EMAIL PROTECTED]>,
"\"Ultima Edizione\"" <[EMAIL PROTECTED]>,
........
a lot of address
On Thu, 17 Jun 1999 19:37:32 +0200 "Alberto Schiappa" wrote:
It doesn't look like a loop to me. It looks like once bounce messsage
(with lots of MIME stuff) being delivered to you.
There are only three Delivered-To: lines in the whole lot, one for the
bounce delivered to you and two for the mail that eventually bounced:
Delivered-To: [EMAIL PROTECTED]
Delivered-To: [EMAIL PROTECTED]
Delivered-To: [EMAIL PROTECTED]
Why do you think it is a loop?
Regards,
Giles
On Wed, 16 Jun 1999 14:18:26 -0500 "Chris Garrigues" wrote:
> We can't very well expect programs that call /usr/lib/sendmail to
> know what to do about a 111 return code, can we? Should that
> progrma mask 111's into 0's?
Possibly. The process of teaching the whole world about two different
sets of mail return codes is painful.
There is a catch though: the behaviour of calling "sendmail" from a
.qmail file would change. Right now scripts that called "sendmail"
from .forward and exited with its exit status can do the same thing
under qmail and "just work".
Teaching the sendmail wrapper to use different exit codes depending on
whether it is running from .qmail isn't a good idea.
I suspect there isn't a "right" answer for all cases.
I've not used dot-forward: how does it handle exit code 75 from
commands that it runs?
Regards,
Giles
> From: Giles Lean <[EMAIL PROTECTED]>
> Date: Thu, 17 Jun 1999 05:49:39 +1000
>
> On Wed, 16 Jun 1999 14:18:26 -0500 "Chris Garrigues" wrote:
>
> > We can't very well expect programs that call /usr/lib/sendmail to
> > know what to do about a 111 return code, can we? Should that
> > progrma mask 111's into 0's?
>
> Possibly. The process of teaching the whole world about two different
> sets of mail return codes is painful.
>
> There is a catch though: the behaviour of calling "sendmail" from a
> .qmail file would change. Right now scripts that called "sendmail"
> from .forward and exited with its exit status can do the same thing
> under qmail and "just work".
I was about to say "Why would you want to call 'sendmail' from a .qmail file,
when you can call qmail-inject?" but I guess your reference to .forward explains
how it might come up. I'd rather that this be special cased in dot-forward
instead, however.
> Teaching the sendmail wrapper to use different exit codes depending on
> whether it is running from .qmail isn't a good idea.
>
> I suspect there isn't a "right" answer for all cases.
>
> I've not used dot-forward: how does it handle exit code 75 from
> commands that it runs?
>
> Regards,
>
> Giles
--
Chris Garrigues virCIO
+1 512 432 4046 4314 Avenue C O-
http://www.DeepEddy.Com/~cwg/ Austin, TX 78751-3709
+1 512 374 0500
My email address is an experiment in SPAM elimination. For an
explanation of what we're doing, see http://www.DeepEddy.Com/tms.html
Nobody ever got fired for buying Microsoft,
but they could get fired for relying on Microsoft.
>> On Tue, 15 Jun 1999 09:48:10 +0300,
>> Anand Buddhdev <[EMAIL PROTECTED]> said:
A> I'm finding that the larger and more complex my qmail setup becomes, the
A> more difficult it is for me to trace the path an address follows through
A> the qmail system.
One thing I always liked about sendmail was using the -d0-99.127 option
to print everything about a delivery in detail.
A> I realise that qmail is modular, and so there's no easy way to determine
A> what really happens. I was thinking of writing a small perl script to
A> "emulate" the decisions made by qmail-send, qmail-lspawn and
A> qmail-rspawn...
I'm more comfortable with the "when in doubt, print it out" approach.
Perhaps having the separate qmail modules check for an environment
variable like QMAILTRACE, and then print where they are and what they're
doing?
--
Karl Vogel
ASC/YCOA, Wright-Patterson AFB, OH 45433, USA
[EMAIL PROTECTED] or [EMAIL PROTECTED]
I have squinted at the cyclog and other man pages in daemontools, but I
don't get the big picture.
I currently use syslog, and rotate the log files once a day, scanning the
logs, making summaries, and keeping a week's worth of compressed daily logs.
cyclog rotates itself when a log reaches a certain size, which is fine, but
what's the right way to collect them and do a daily scan and crunch?
Regards,
John Levine, [EMAIL PROTECTED], Primary Perpetrator of "The Internet for Dummies",
Information Superhighwayman wanna-be, http://iecc.com/johnl, Sewer Commissioner
Finger for PGP key, f'print = 3A 5B D0 3F D9 A0 6A A4 2D AC 1E 9E A6 36 A3 47
John R Levine writes:
> I have squinted at the cyclog and other man pages in daemontools, but I
> don't get the big picture.
>
> I currently use syslog, and rotate the log files once a day, scanning the
> logs, making summaries, and keeping a week's worth of compressed daily logs.
>
> cyclog rotates itself when a log reaches a certain size, which is fine, but
> what's the right way to collect them and do a daily scan and crunch?
Unfortunately, cyclog currently has no way to tell you when it's
rotated a log. That's too bad, because you have to poll. It's
definitely abandoned a log when it removes write permissions, though,
so you could configure it to write a very large number of files of
relatively small size, and once a day, concatenate all the read-only
files, grovel through them, and compress them.
--
-russ nelson <[EMAIL PROTECTED]> http://crynwr.com/~nelson
Crynwr supports Open Source(tm) Software| PGPok | Government schools are so
521 Pleasant Valley Rd. | +1 315 268 1925 voice | bad that any rank amateur
Potsdam, NY 13676-3213 | +1 315 268 9201 FAX | can outdo them. Homeschool!
On Thu, Jun 17, 1999 at 03:01:54PM -0400, John R Levine wrote:
> I have squinted at the cyclog and other man pages in daemontools, but I
> don't get the big picture.
>
> I currently use syslog, and rotate the log files once a day, scanning the
> logs, making summaries, and keeping a week's worth of compressed daily logs.
>
> cyclog rotates itself when a log reaches a certain size, which is fine, but
> what's the right way to collect them and do a daily scan and crunch?
Bruce Guenter has written a patch for cyclog that allows one to specify a
program to run when cyclog is about to rotate the logs. You could then have
a little shell script which mv's the oldest log to an archive area, where
you can then analyse it at leisure. I use it myself, and I like it. Find
the patch at:
http://www.qcc.sk.ca/~bguenter/distrib/daemontools
I hope Dan will incorporate this functionality into the main daemontools
distribution.
--
Anand
<<< multipart/signed; boundary=WfZ7S8PLGjBY9Voh; micalg=pgp-md5;protocol="application/pgp-signature": Unrecognized >>>
On Thu, 17 Jun 1999 [EMAIL PROTECTED] wrote:
> Hello,
>
> I've got a quick question. Is it safe to kill qmail? I'm not talking about
> kill -9 or anything, just a normal kill.
yes.
> The thing is that since there doesn't seem to be a way to get qmail to
> reread it's configuration files, I had to use a small kill-and-restart
it re-reads many of the configuration values -- read the documenation
about which ones it does handle properly
Richard
On Thursday, June 17, 1999 4:23 PM, Richard Letts
[SMTP:[EMAIL PROTECTED]] wrote:
> > The thing is that since there doesn't seem to be a way to get qmail to
> > reread it's configuration files, I had to use a small kill-and-restart
> it re-reads many of the configuration values -- read the documenation
> about which ones it does handle properly
Actually, the docs are rather quiet on the way *most* of the config files
are reread.
What is clear, however, is that "locals" and "virtualdomains" are both
reread by HUPing "qmail-send". And from experience, I also know that
"rcpthosts" and "badmailfrom" are reread on the fly.
My impression is that, except for what is documented to the contrary, most
files are reread on the fly.
HTH.
Dave
+ Dave Kitabjian <[EMAIL PROTECTED]>:
| And from experience, I also know that
| "rcpthosts" and "badmailfrom" are reread on the fly.
This is because they are read by qmail-smtpd, and a new copy of that
program is run on each incoming SMTP connection.
| My impression is that, except for what is documented to the
| contrary, most files are reread on the fly.
Only those read by programs that, like qmail-smtpd, are being run
afresh on demand: qmail-remote, qmail-inject, qmail-qmqpc. Which
leaves only qmail-send; see the qmail-control manual page. So yes,
your impression is correct, since the qmail-send manual page contains
an explicit warning about this.
- Harald
Heinz Wittenbecher wrote:
>
> Apologies for the slightly off topic, but I suspect that I'm not the only
> qmail/cyrus user.
Okay, I'm even more off-topic.
> My problem is that I'm unable to share user folders and folders created with
> the cyrus create mailfolder command as folder instead of user.folder are
> visible to everyone. I'm primarily using IE 4 and 5 as clients. Netscape is
> worse. It shows all the folders, even other users'.
If I understand exactly what you're saying here, you can share user
folders by adding access lists to the folders. I don't remember the
exact syntax, but it's going to be something like
sam everyone lrc user.*
Now for the caveats and notes: sam is short for setaclmailbox. Everyone
is the user you're setting access for, but I'm not positive that it's
everyone and not everybody or all. lrc is a string of permission bits,
but I can't remember which ones you really need, and it really depends
on what you want other users to be able to do with other users'
mailboxes. user.* means all inboxes. You definitely want to create
user mailboxes under user, not in the root.
Oh, yeah, and I'm not certain I have those args in the right order. :-|
Now, if anyone could remind me where I put that man page...
Or you could just use the gui configuration utility provided cyradm.
--Dave
At 04:20 PM 6/17/99 , Eric Shafto:
>Heinz Wittenbecher wrote:
> >
> > Apologies for the slightly off topic, but I suspect that I'm not the only
> > qmail/cyrus user.
>
> > My problem is that I'm unable to share user folders and folders created
> with
> > the cyrus create mailfolder command as folder instead of user.folder are
> > visible to everyone. I'm primarily using IE 4 and 5 as clients. Netscape is
> > worse. It shows all the folders, even other users'.
>
>If I understand exactly what you're saying here, you can share user
>folders by adding access lists to the folders. I don't remember the
>exact syntax, but it's going to be something like
>
>sam everyone lrc user.*
Yesterday I had just finished getting qmail running, and I wanted to place
the command to start qmail in my boot script. I am using Mandrake, and I
*thought* I found the bootup script under rc.sysinit, so I placed the qmail
start line in that script. Well, when I rebooted, all hell broke loose
when lilo started it's turning on of services. There were so many error
messages that I couldn't keep up with them, so I just decided to nuke and
pave.
I don't wish to have this problem again.. so does anyone know where the
boot script is for Mandrake version of RedHat 6? And, what I should *not*
do when placing a new line in that script? Thanks for any help.
james
There are 2 ways to do it ( I believe ).
The simpler method is to add a single line into /etc/rc.local
Add a line that calls your qmail start up script.
Matt.
> -----Original Message-----
> From: James [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 8:23 AM
> To: [EMAIL PROTECTED]
> Subject: Boot file
>
> Yesterday I had just finished getting qmail running, and I wanted to
> place
> the command to start qmail in my boot script. I am using Mandrake,
> and I
> *thought* I found the bootup script under rc.sysinit, so I placed the
> qmail
> start line in that script. Well, when I rebooted, all hell broke
> loose
> when lilo started it's turning on of services. There were so many
> error
> messages that I couldn't keep up with them, so I just decided to nuke
> and
> pave.
>
> I don't wish to have this problem again.. so does anyone know where
> the
> boot script is for Mandrake version of RedHat 6? And, what I should
> *not*
> do when placing a new line in that script? Thanks for any help.
>
> james
I placed mine in my /etc/rc.local file
I forget the exact line but it was the one listed in the install file
something like:
csh -cf '/var/qmail/rd &'
I THINK that's what it was, you should check.
At any rate, as near as I can tell, /etc/rc.local is where the autostart
things ought to go.
It seems to launch just fine as if I had launched it by hand and passes
TEST.deliver
I have exactly the same system as you, Linux Mandrake.
Where I am having trouble is in getting the SMTP host to work. The step for
commenting out a line in the /etc/rc.d/init.d/sendmail
The expected line is not there but a lot of code that seems to build up that
line is there.
I'd also be curious how the version of MUTT that came with Mandrake behaves.
It seems to prefer Mailboxes to Maildirs.
Alex Miller
> -----Original Message-----
> From: James [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, June 17, 1999 8:23 AM
> To: [EMAIL PROTECTED]
> Subject: Boot file
>
>
> Yesterday I had just finished getting qmail running, and I wanted to place
> the command to start qmail in my boot script. I am using Mandrake, and I
> *thought* I found the bootup script under rc.sysinit, so I placed
> the qmail
> start line in that script. Well, when I rebooted, all hell broke loose
> when lilo started it's turning on of services. There were so many error
> messages that I couldn't keep up with them, so I just decided to nuke and
> pave.
>
> I don't wish to have this problem again.. so does anyone know where the
> boot script is for Mandrake version of RedHat 6? And, what I should *not*
> do when placing a new line in that script? Thanks for any help.
>
> james
>
>
Several people on the list have reported timeout problems
with clients like Outlook when using tcpserver and qmail's pop3.
Unlike many replies suggest, it isn't a problem with client
software. It's unlikely to be a problem with qmail's pop3
server either.
I just got a note from our tech support department that people
were intermittantly having problems downloading large
attachemnts, starting a couple days ago. They wanted to know
if it was the "new mail server".
I dismissed this off hand since I hadn't installed qmail yet,
but then I realized I had installed a tiny little bit 'o qmail
a couple days ago. I replaced our cucipop in daemon mode
with cucipop running under tcpserver.
Something is wrong with tcpserver.
--
Aaron Nabil
Aaron Nabil writes:
>
> Several people on the list have reported timeout problems
> with clients like Outlook when using tcpserver and qmail's pop3.
>
> Unlike many replies suggest, it isn't a problem with client
> software. It's unlikely to be a problem with qmail's pop3
> server either.
>
> I just got a note from our tech support department that people
> were intermittantly having problems downloading large
> attachemnts, starting a couple days ago. They wanted to know
> if it was the "new mail server".
>
> I dismissed this off hand since I hadn't installed qmail yet,
> but then I realized I had installed a tiny little bit 'o qmail
> a couple days ago. I replaced our cucipop in daemon mode
> with cucipop running under tcpserver.
>
> Something is wrong with tcpserver.
Before you start making conclusions, try to actually look at tcpserver's
source code.
After the connection is received, tcpserver forks and execs the server, at
which point the forked tcpserver ceases to exist.
It would be rather impressive for tcpserver to affect an existing
connection when it is no longer running.
--
Sam
Sam writes...
>Aaron Nabil writes:
>
>>
>> Several people on the list have reported timeout problems
>> with clients like Outlook when using tcpserver and qmail's pop3.
>>
>> . . .
>>
>> Something is wrong with tcpserver.
>
>Before you start making conclusions, try to actually look at tcpserver's
>source code.
>
>After the connection is received, tcpserver forks and execs the server, at
>which point the forked tcpserver ceases to exist.
>
>It would be rather impressive for tcpserver to affect an existing
>connection when it is no longer running.
Tcpserver sets up the environment, signal handling and masks, and
creates the descriptors (and sets options on them) for the child. It
_would_ be impressive for tcpserver to affect a forked and execed pop
connection, but that's hardly germaine. To bugger up a pop session,
all the tcpserver has to do is pollute the signal environment, fail to
set or clear some critical process limit, or mess up one of the
descriptors it passed to the child in any one of countless ways. It
would be a trivial patch to tcpserver to cause your popserver
sessions to magically self-destruct after being connected for 30
seconds (not withstanding that it would be "impressive for tcpserver
to affect an existing connection when it is no longer running"), long
after tcpserver has forked and exec'd itself in oblivion.
The only two things I've noticed about tcpserver so far are that
this setsockopt call seems kinda risky (from having seen a few
setsockopt implementations, I wouldn't be surprised if the call fails).
if (flagkillopts) {
setsockopt(t,IPPROTO_IP,1,(char *) 0,0); /* 1 == IP_OPTIONS */
/* if it fails, bummer */
}
I would have used the safer and undoubtably correct...
if (flagkillopts) {
int opt = 0;
setsockopt(t,IPPROTO_IP,1,&opt,sizeof(opt)); /* 1 == IP_OPTIONS */
/* if it fails, bummer */
}
And that tcpserver, unlike inetd, doesn't attach stderr to the
network connection. In fact all of the children inherit tcpserver's
stderr! I don't see any obvious side-effect of this yet.
--
Aaron Nabil
+ Aaron Nabil <[EMAIL PROTECTED]>:
| And that tcpserver, unlike inetd, doesn't attach stderr to the
| network connection. In fact all of the children inherit tcpserver's
| stderr! I don't see any obvious side-effect of this yet.
This is undoubtedly intentional, and quite useful. You can just pipe
the stderr of tcpserver into your favourite logging tool, and have the
spawned children log stuff that way. If you ever *need* to have the
child's stderr connected to the network socket, it is of course
trivial to arrange this.
- Harald
Dear Sir:
My boss wants all messages out to be appended with one or two lines
of 'trailer' (footer), I think I'm familiar with qmail, but I can't
find a suitable place to do so. (something like what hotmail does)
Could anybody give me some advice? I might need to modify the code
myself, that's OK, but I'm not sure which one is better.
qmail-queue? qmail-smtpd? qmail-send? qmail-remote?
I have to make sure inner message transfers will not cause "doubled
-added", that's what I found if I modified the qmail-queue.
Thanks for your help.
--
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~+
| Shan-Ta CHEN E-Mail : [EMAIL PROTECTED] |
| Silver CHEN Tel(O) : +886-2-2773-9858-288 |
| �����F Tel(H) : +886-2-2914-1402 |
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~+
