qmail Digest 17 Apr 1999 10:00:01 -0000 Issue 613
Topics (messages 24338 through 24394):
Foreign language and qmail bounces
24338 by: [EMAIL PROTECTED]
24356 by: "Roman V. Isaev" <[EMAIL PROTECTED]>
24373 by: "Fred Lindberg" <[EMAIL PROTECTED]>
Help with qmail and RedHat 5.2!!
24339 by: "Henrik Holmberg" <[EMAIL PROTECTED]>
24340 by: Bart Blanquart <[EMAIL PROTECTED]>
aliases with a dash (-)
24341 by: Marlon Anthony Abao <[EMAIL PROTECTED]>
FW: Web Interface to Qmail on Linux
24342 by: Tom Hukins <[EMAIL PROTECTED]>
Help...Pleeeeaaaase! qmail-smtpd and diald
24343 by: "Joerg Toellner" <[EMAIL PROTECTED]>
It still cannot deliver local messages...
24344 by: Jim Beam <[EMAIL PROTECTED]>
24362 by: Harald Hanche-Olsen <[EMAIL PROTECTED]>
24379 by: Jos Backus <[EMAIL PROTECTED]>
small problem
24345 by: "Jim Baxter" <[EMAIL PROTECTED]>
24346 by: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
24349 by: "Adam D. McKenna" <[EMAIL PROTECTED]>
24350 by: "Adam D. McKenna" <[EMAIL PROTECTED]>
24351 by: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
24352 by: "Adam D. McKenna" <[EMAIL PROTECTED]>
24353 by: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
24363 by: "Jay D. Dyson" <[EMAIL PROTECTED]>
24368 by: Troy Morrison <[EMAIL PROTECTED]>
24370 by: "Scott Burkhalter" <[EMAIL PROTECTED]>
Forward using autoturn?
24347 by: "Ramon Oliver" <[EMAIL PROTECTED]>
MICROSOFT'S HOTMAIL USES QMAIL!!!!
24348 by: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
hardware
24354 by: "A.Wadas" <[EMAIL PROTECTED]>
24357 by: Chris Johnson <[EMAIL PROTECTED]>
24361 by: "A.Wadas" <[EMAIL PROTECTED]>
24366 by: David A Galbraith CIRT <[EMAIL PROTECTED]>
24369 by: "A.Wadas" <[EMAIL PROTECTED]>
24372 by: "Sam" <[EMAIL PROTECTED]>
Army 1, qmail 0
24355 by: "Fred Lindberg" <[EMAIL PROTECTED]>
24365 by: Russell Nelson <[EMAIL PROTECTED]>
feedback when message cannot be delivered immediately
24358 by: Eric Dahnke <[EMAIL PROTECTED]>
Thanks for the help...
24359 by: Jim Beam <[EMAIL PROTECTED]>
Another q about open relaying
24360 by: Stanley Horwitz <[EMAIL PROTECTED]>
24377 by: Chris Johnson <[EMAIL PROTECTED]>
www.qmail.com appears to have nothing to do with qmail
24364 by: Mark Delany <[EMAIL PROTECTED]>
Dashes in qmail aliasing.
24367 by: "Jay D. Dyson" <[EMAIL PROTECTED]>
24393 by: Marlon Anthony Abao <[EMAIL PROTECTED]>
I feel really dumb now...
24371 by: Jim Beam <[EMAIL PROTECTED]>
SMTP Delivery Problem from Outlook to Qmail
24374 by: "Robert Harrison" <[EMAIL PROTECTED]>
24376 by: Chris Johnson <[EMAIL PROTECTED]>
24394 by: "Robert Harrison" <[EMAIL PROTECTED]>
smtp question
24375 by: Rob Genovesi <[EMAIL PROTECTED]>
24380 by: John Grant <[EMAIL PROTECTED]>
qmail & listserv
24378 by: "Swanson, Scott" <[EMAIL PROTECTED]>
24381 by: David Villeger <[EMAIL PROTECTED]>
qmqpc reliability
24382 by: "Robin Bowes" <[EMAIL PROTECTED]>
24383 by: Mark Delany <[EMAIL PROTECTED]>
24384 by: "Robin Bowes" <[EMAIL PROTECTED]>
24385 by: Mark Delany <[EMAIL PROTECTED]>
Solaris 7 x86 and number of process limitation
24386 by: [EMAIL PROTECTED] ()
24387 by: Mark Delany <[EMAIL PROTECTED]>
24389 by: "Sam" <[EMAIL PROTECTED]>
24391 by: Mark Delany <[EMAIL PROTECTED]>
extra fields in users/assign?
24388 by: "Theodore Cekan" <[EMAIL PROTECTED]>
24390 by: Chris Johnson <[EMAIL PROTECTED]>
qmail delivery to Exchange
24392 by: Scott Burkhalter <sbur@[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]
----------------------------------------------------------------------
>
> I wouldn't do it. It would help your users but confuse more people around
> the world that cannot understand your localized bounce messages. Bounces
> typically go to outside users.
>
> In any case your users still get bounce messages from mailers all over the
> world in English.
>
> Regards, Frank
Hi Frank,
I didn't explain properly. I won't replace English
descriptions but I will add locale descriptions. Futhermore, our bounces also are
delivered to our internal users when our qmail gateway detects a failure condition in
its connections (ie. unknown host). Nevertheless, your hint about remote bounces is
really
useful. I have forgot this case. If our internal users
will receive remote bounce messages in English, maybe it isn't worthwhile to translate
ours.
However also I have thinked to explain RFC1893 failure
codes (as #4.1.2) in a our intranet, but it seems to me
that qmail produces the same code with different
descriptions. Am I right?
David.
====
David Jorrin <[EMAIL PROTECTED]>
"This chapter is about Laziness, Impatience
and Hubris because this chapter is about
good software design"
Larry Wall, Tom Christiansen &
Randal L. Schwartz [Programming Perl]
----------------------------------------------------------------
Get your free email from AltaVista at http://altavista.iname.com
Hello Frank,
Friday, April 16, 1999, 12:28:22 PM, you wrote:
>> Do you have any suggestion or solution?
FT> I wouldn't do it. It would help your users but confuse more people around
FT> the world that cannot understand your localized bounce messages. Bounces
FT> typically go to outside users.
FT> In any case your users still get bounce messages from mailers all over the
FT> world in English.
It's kinda difficult to explain 'message bounce' to some stupid clients.
So _adding_ other language text to the bounce is a good thing... I'd like this
feature, but only if it will add "Content-Type: text/plain; charset=koi8-r" to
the header as well -- otherwise it will be nearly useless, at least with Russian
(hell,there are 6 enconding types).
--
Roman V. Isaev http://www.gunlab.com.ru Moscow, Russia
On Fri, 16 Apr 1999 06:04:24 -0400 (EDT), [EMAIL PROTECTED]
wrote:
>However also I have thinked to explain RFC1893 failure
>codes (as #4.1.2) in a our intranet, but it seems to me
>that qmail produces the same code with different
>descriptions. Am I right?
Yes, 4.1.2 is a machine-readable code per rfc1893. The English "human
readable" code is more descriptive. For instance:
"Message rejected because you are not a moderator" and "Message
rejected because moderation commands are not allowed for this list" may
lead to the same code, but convey different information to the reader.
You could however, map the codes to Russian and make it "Message
rejected: insufficient privilege". Especially when you add it to a
English description, this should work well.
-Sincerely, Fred
(Frederik Lindberg, Infectious Diseases, WashU, St. Louis, MO, USA)
Hi I tried to install qmail on my redhat and after that no mail
worked :( What did I do wrong???
I used this ftp://rpmfind.net/linux/solaris/RPMS/i386//qmail-1.01-
8.i386.rpm
and I followed the INSTALL.redhat and so on but it didn't work. I
really need help because I don't want to run sendmail the VERY
insecure MTA!
/Henrik
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
Henrik Holmberg wrote:
>
> Hi I tried to install qmail on my redhat and after that no mail
> worked :( What did I do wrong???
>
You did remove sendmail?
and
_what_ isn't working? What do the logs say?
> I used this ftp://rpmfind.net/linux/solaris/RPMS/i386//qmail-1.01-
> 8.i386.rpm
There is a newer version of qmail. You might want to install that instead.
bt
--
Bart Blanquart
[EMAIL PROTECTED]
tel (02)50 51 916 fax (02)50 51 930
"There are only two industries that refer to their customers as 'users'." --Edward
Tufte
hello,
is there something wrong with an alias with a dash (-)? i can't seem to
compile my aliases file with some alias that contains a dash in it.
any help would really be appreciated.
-marlon
On Thu, Apr 15, 1999 at 10:25:26PM +0000, Sam wrote:
> Tom Hukins writes:
>
> > Sqwebmail also refuses logins through proxy farms and proxy
> > hierarchies.
>
> Only broken proxies.
Not true. Proxy farms and hierarchies are not broken. You have yet
to provide any reason other than prejudice to explain why they are
broken. I have cited the HTTP specification to explain why you are
wrong.
> I have no problems logging in from behind my company's
> proxy/firewall.
I assume that's because it's just the one proxy always using the
same IP address.
According to the HTTP specification, Sqwebmail is broken. I think
this is a shame, as otherwise it's a great product.
I know this isn't on-topic for the qmail list, but I thought I
should warn anyone who might want to run Sqwebmail. A full
explanation can be found in the Sqwebmail list archives. I won't
mention this on the qmail list any more.
Regards,
Tom
Hi all Qmail-/Linux-/Email-/diald-Gods,
i have a strange problem. I think it isnt concerned right to qmail (sorry
for a little bit offtopic) but to my email sending/receiving. I hope someone
of you can maybe enlighten me.
The story:
----------
I´ve set up a mailserver under S.u.S.E. Linux 6.0 and use Qmail 1.03 as my
MTA. As i use a dialup ppp-connection to my ISP i wish my linux-commserver
as a relay for my outgoing email (to all of the world) as well as a
local-delivery-system for inhouse communication. As workstations we use
Win95/98 computers and outlook as MUA.
To bring up the dialup-connection (not even for email but also for WWW
a.s.o) i use diald and its rules. This works fine.
i installed and setup qmail and my outlook. Local mails go through the linux
server to home-maildirs of local users and outgoing mails were collected via
a catchall-alias in /var/qmail/control/virtualdomains in a single maildir
where cron shoots them out regularly using sendmail. Incoming mails were
fetched via fetchmail and delivered locally. So far so good...all components
do their jobs very well and fast. BUT...
Problem:
--------
Just in the moment i start diald (the dialer-daemon) the server didnt accept
my mails from my win95-client with outlook further. The new mail waits in
the outbound-folder for delivering in my outlook. No error-window appears.
Waiting for hours and hours. Nothing happens. The /var/log/mail file shows
nothing about the new mail (it does so when diald is not running and the
mail delivery works).
If i kill the diald-process and it didnt appear any longer in process list
the waiting mail in my outlook vanishes from my outbound-folder and appears
on the linux server in the correct place (i.E. in a local-user-homedir or in
the collect-the-relaymail-maildir) - all works fine again.
I changed NOTHING in the configurations. Neither in qmail/linux side nor at
the Win/outlook side. The only thing i do is to start diald (no mail
function further) and kill diald (back to normal function) via a telnet
session, nothing else.
Im helpless. I cant imagine what there can be interfere each other. On
another machine (my test computer) i set up the same configuration for
testing and there it works fine with or without diald running. So there must
be a way to solve this problem and it seems only a configuration problem (in
diald???) on the other machine. But i cant figure it out. Compared all
related files on both machines for hours...i cant see any difference.
Anyone can give me a clue what to look for or check out? Any hint
appreciated...really!
If you need more Informations (like excerpts from my diald.conf-file or so)
please dont hesitate to contact me for it. Ill do (almost) anything to get
help and solve this riddle.
I will be so thankful for your soon replies here in the list or via email
to: [EMAIL PROTECTED] Thx. in advance and sorry for offtopic again. I hope i
bothered you not too much.
Here is the error I am getting:
Apr 16 06:39:50 smtp qmail: 924262790.544729 starting delivery 8: msg
1214814 to local [EMAIL PROTECTED]
Apr 16 06:39:50 smtp qmail: 924262790.544934 status: local 1/10 remote
0/20
Apr 16 06:39:50 smtp qmail: 924262790.561974 delivery 8: deferral:
Unable_to_chdir_to_maildir._(#4.2.1)/
Apr 16 06:39:50 smtp qmail: 924262790.562174 status: local 0/10 remote
0/20
Apr 16 06:45:55 smtp qmail: 924263155.564732 starting delivery 9: msg
1214813 to local [EMAIL PROTECTED]
Apr 16 06:45:55 smtp qmail: 924263155.564938 status: local 1/10 remote
0/20
Apr 16 06:45:55 smtp qmail: 924263155.581988 delivery 9: deferral:
Unable_to_chdir_to_maildir._(#4.2.1)/
Apr 16 06:45:55 smtp qmail: 924263155.582178 status: local 0/10 remote
0/20
I have the /Maildir/ entry correct, and have ./Maildir/ in my .qmail
file in the /home/pop - what can I be missing - I dont think it is a
permissions thing.
Thanks again - in advance.
------------------------------------------------------------------------
----------------
James Beam
Support
PDQ.net ENTECH.com
http://www.pdq.net http://www.entech.com
------------------------------------------------------------------------
----------------
<<Jim Beam.vcf>>
Jim Beam.vcf
+ Jim Beam <[EMAIL PROTECTED]>:
| Apr 16 06:39:50 smtp qmail: 924262790.544729 starting delivery 8: msg
| 1214814 to local [EMAIL PROTECTED]
| Apr 16 06:39:50 smtp qmail: 924262790.561974 delivery 8: deferral:
| Unable_to_chdir_to_maildir._(#4.2.1)/
|
| I have the /Maildir/ entry correct, and have ./Maildir/ in my .qmail
| file in the /home/pop - what can I be missing - I dont think it is a
| permissions thing.
Why don't you think so? Does user pop own /home/pop/Maildir/ or not?
The message you get means what it says; if you think permissions are
OK, you should give us some evidence to refute, or we can do nothing
to help.
- Harald
On Fri, Apr 16, 1999 at 07:56:40AM -0500, Jim Beam wrote:
> Apr 16 06:45:55 smtp qmail: 924263155.581988 delivery 9: deferral:
> Unable_to_chdir_to_maildir._(#4.2.1)/
The fastest way to fix this is probably to ktrace/truss qmail-lspawn and its
children, kill -ALRM qmail-send and check the syscall tracelog for the reason
for the chdir failing.
--
Jos Backus _/ _/_/_/ "Reliability means never
_/ _/ _/ having to say you're sorry."
_/ _/_/_/ -- D. J. Bernstein
_/ _/ _/ _/
[EMAIL PROTECTED] _/_/ _/_/_/ use Std::Disclaimer;
Hi.,
I am having a small little problem now that i've setup qmail.. actually,
i don't know if it existed before because no mail was ever transferred on my
system before setting up qmail. anyways, every time i download and delete my
mail via POP, a message from the MailerDaemon keeps being sent to my mail
box
--------
This text is part of the internal format of your mail folder, and is not
a real message. It is created automatically by the mail system software.
If deleted, important folder data will be lost, and it will be re-created
with the data reset to initial values.
--------
that would be the text in the message.. i have no idea where it is coming
from, can anyone speculate?
thanks
Jim
[EMAIL PROTECTED]
AFAIK this is a quirk of the mbox format. You could migrate to Maildir
format (depending on which MUA your Unix users use) or simply ignore
these messages and go on with life.
If everybody reads messages via POP, you can use Maildir and qmail's pop
server and everyone will be happy. Be sure to read the FAQ at qmail.org
beforehand!
Jim Baxter wrote:
>
> Hi.,
>
> I am having a small little problem now that i've setup qmail.. actually,
> i don't know if it existed before because no mail was ever transferred on my
> system before setting up qmail. anyways, every time i download and delete my
> mail via POP, a message from the MailerDaemon keeps being sent to my mail
> box
>
> --------
> This text is part of the internal format of your mail folder, and is not
> a real message. It is created automatically by the mail system software.
> If deleted, important folder data will be lost, and it will be re-created
> with the data reset to initial values.
> --------
>
> that would be the text in the message.. i have no idea where it is coming
> from, can anyone speculate?
--
___THE___ "Commercial OS vendors are, at the moment, all closed
\ \ / / economies, and doomed to fall in their competition with
\ V / open economies just as communism eventually fell."
\ / -- H. Reiser, Unix OS developer
/ \ _____________________________________________________
/ ^ \ | Juan Carlos Castro y Castro - [EMAIL PROTECTED] |
/ / \ \ | Diretor de Informática e Eventos Sobrenaturais da |
~~~ ~~~ | E-RACE CORPORATION |
RACER -----------------------------------------------------
From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
: AFAIK this is a quirk of the mbox format. You could migrate to Maildir
: format (depending on which MUA your Unix users use) or simply ignore
: these messages and go on with life.
From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
: AFAIK this is a quirk of the mbox format. You could migrate to Maildir
: format (depending on which MUA your Unix users use) or simply ignore
: these messages and go on with life.
Er, sorry about that blank message.
Anyway, I am pretty sure this has to do with PINE and not qmail in any way.
--Adam
"Adam D. McKenna" wrote:
>
> From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
>
> : AFAIK this is a quirk of the mbox format. You could migrate to Maildir
> : format (depending on which MUA your Unix users use) or simply ignore
> : these messages and go on with life.
>
> I am pretty sure this has to do with PINE and not qmail in any way.
Isn't PINE one of those which don't read Maildirs? I can't see how a MUA
that reads via POP could stuff these bogus messages on a mailbox.
--
___THE___ "Commercial OS vendors are, at the moment, all closed
\ \ / / economies, and doomed to fall in their competition with
\ V / open economies just as communism eventually fell."
\ / -- H. Reiser, Unix OS developer
/ \ _____________________________________________________
/ ^ \ | Juan Carlos Castro y Castro - [EMAIL PROTECTED] |
/ / \ \ | Diretor de Informática e Eventos Sobrenaturais da |
~~~ ~~~ | E-RACE CORPORATION |
RACER -----------------------------------------------------
From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
: "Adam D. McKenna" wrote:
: >
: > From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
: >
: > : AFAIK this is a quirk of the mbox format. You could migrate to Maildir
: > : format (depending on which MUA your Unix users use) or simply ignore
: > : these messages and go on with life.
: >
: > I am pretty sure this has to do with PINE and not qmail in any way.
:
: Isn't PINE one of those which don't read Maildirs? I can't see how a MUA
: that reads via POP could stuff these bogus messages on a mailbox.
Well whatever it is, it's not qmail. I used cucipop with a mbox format for
quite a while with qmail and never got any messages like this. It's most
likely the POP daemon doing this.
--Adam
"Adam D. McKenna" wrote:
>
> From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
>
> : "Adam D. McKenna" wrote:
> : >
> : > From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
> : >
> : > : AFAIK this is a quirk of the mbox format. You could migrate to Maildir
> : > : format (depending on which MUA your Unix users use) or simply ignore
> : > : these messages and go on with life.
> : >
> : > I am pretty sure this has to do with PINE and not qmail in any way.
> :
> : Isn't PINE one of those which don't read Maildirs? I can't see how a MUA
> : that reads via POP could stuff these bogus messages on a mailbox.
>
> Well whatever it is, it's not qmail. I used cucipop with a mbox format for
> quite a while with qmail and never got any messages like this. It's most
> likely the POP daemon doing this.
Bingo! That should be the cause. So using a different pop server (for
instance qmail's own) will solve Mr. Baxter's problem. And switching to
Maildir is healthy anyway.
--
___THE___ "Commercial OS vendors are, at the moment, all closed
\ \ / / economies, and doomed to fall in their competition with
\ V / open economies just as communism eventually fell."
\ / -- H. Reiser, Unix OS developer
/ \ _____________________________________________________
/ ^ \ | Juan Carlos Castro y Castro - [EMAIL PROTECTED] |
/ / \ \ | Diretor de Informática e Eventos Sobrenaturais da |
~~~ ~~~ | E-RACE CORPORATION |
RACER -----------------------------------------------------
-----BEGIN PGP SIGNED MESSAGE-----
On Fri, 16 Apr 1999, Juan Carlos Castro y Castro wrote:
> > I am pretty sure this has to do with PINE and not qmail in any way.
>
> Isn't PINE one of those which don't read Maildirs? I can't see how a MUA
> that reads via POP could stuff these bogus messages on a mailbox.
It's PINE all right, and it does the job via IMAP. Personally, I
find the whole thing annoying, so I haven't upgraded from v3.96. It ticks
me off that I can have a totally empty mailbox and still get the "You have
mail" when I login. Of course, I check the mail and there's nothing
there, but that's a "feature," not a bug, right? Riiiiight.
- -Jay
( ______
)) .-- "There's always time for a good cup of coffee." --. >===<--.
C|~~| (>-- Jay D. Dyson -- [EMAIL PROTECTED] --<) | = |-'
`--' `-- People who think NASA is fake view WWF as real. --' `-----'
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
iQCVAwUBNxdYD82OVDpaKXD9AQGO5AP8CSsiETm5Eq9lUaOpBU24yo8hhWTI3ueo
F8e0m5/CbOlSJDGmO9b7mH9lzjBnkgvSz9hB89BLi8bNDzsbTR+Q5mfeHGDNoTmY
i2HbGKy0ndTIFOjn1iMOcj7xExmyOGS0zpfbondIuLzujSp6U1gki4o5P4us45Dq
CeGzaeBJrn0=
=FYwE
-----END PGP SIGNATURE-----
| Well whatever it is, it's not qmail. I used cucipop with a mbox format for
| quite a while with qmail and never got any messages like this. It's most
| likely the POP daemon doing this.
You're right. It's Pine, *and* is the UW ipop daemon that comes with
Pine, *and* it's the UW imap daemon that comes with pine (or is available
separately). In Pine, there is a "quell-folder-internal-msg" option that
can turn this off.
Hmm... I just went to confirm that the IW ipop daemon does this too, and I
can't find it, so I might be wrong about that. I know that Pine/imapd do
generate it, though.
Troy
I get this same AFAIK activity on a system that uses Sendmail when I
retrieve my mail through MS Outlook through POP3. It must be the POP
daemon.
Scott Burkhalter
VP Engineering & Technology
Entyre Doc Prep, Inc.
[EMAIL PROTECTED]
-----Original Message-----
From: Juan Carlos Castro y Castro [mailto:[EMAIL PROTECTED]]
Sent: Friday, April 16, 1999 10:47 AM
To: Qmail List
Cc: Adam D. McKenna; Jim Baxter
Subject: Re: small problem
"Adam D. McKenna" wrote:
>
> From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
>
> : "Adam D. McKenna" wrote:
> : >
> : > From: Juan Carlos Castro y Castro <[EMAIL PROTECTED]>
> : >
> : > : AFAIK this is a quirk of the mbox format. You could migrate to
Maildir
> : > : format (depending on which MUA your Unix users use) or simply ignore
> : > : these messages and go on with life.
> : >
> : > I am pretty sure this has to do with PINE and not qmail in any way.
> :
> : Isn't PINE one of those which don't read Maildirs? I can't see how a MUA
> : that reads via POP could stuff these bogus messages on a mailbox.
>
> Well whatever it is, it's not qmail. I used cucipop with a mbox format
for
> quite a while with qmail and never got any messages like this. It's most
> likely the POP daemon doing this.
Bingo! That should be the cause. So using a different pop server (for
instance qmail's own) will solve Mr. Baxter's problem. And switching to
Maildir is healthy anyway.
--
___THE___ "Commercial OS vendors are, at the moment, all closed
\ \ / / economies, and doomed to fall in their competition with
\ V / open economies just as communism eventually fell."
\ / -- H. Reiser, Unix OS developer
/ \ _____________________________________________________
/ ^ \ | Juan Carlos Castro y Castro - [EMAIL PROTECTED] |
/ / \ \ | Diretor de Informática e Eventos Sobrenaturais da |
~~~ ~~~ | E-RACE CORPORATION |
RACER -----------------------------------------------------
Hi everybody,
I've implemented qmail with the autoturn option,
but I'd
like now to forward all incoming messages for a
certain
domain, domain.com, to a single e-mail
address,
Thanks in advance,
Carles.
I already knew (as everybody) that MS couldn't put NT to work properly
and uses Solaris to run HotMail. But this is new. Or not. Forgive me if
this is old news.
Try to send mail to [EMAIL PROTECTED] and
see just WHAT bounces back! ;)
--
___THE___ "Commercial OS vendors are, at the moment, all closed
\ \ / / economies, and doomed to fall in their competition with
\ V / open economies just as communism eventually fell."
\ / -- H. Reiser, Unix OS developer
/ \ _____________________________________________________
/ ^ \ | Juan Carlos Castro y Castro - [EMAIL PROTECTED] |
/ / \ \ | Diretor de Informática e Eventos Sobrenaturais da |
~~~ ~~~ | E-RACE CORPORATION |
RACER -----------------------------------------------------
Running Qmail on UltraWide disk and having at the moment no problems
makes me thinking
how to save money. Does it hurt significantly performance If I change
the disk to IDE one
Qmail is doing not more than 300 messages a day .
Appreciate a comment.
Andrzej
On Fri, Apr 16, 1999 at 04:46:37PM +0000, A.Wadas wrote:
> Running Qmail on UltraWide disk and having at the moment no problems
> makes me thinking
> how to save money. Does it hurt significantly performance If I change
> the disk to IDE one
> Qmail is doing not more than 300 messages a day .
You could do 300 messages a day off of a floppy disk.
Chris
Chris Johnson wrote:
> On Fri, Apr 16, 1999 at 04:46:37PM +0000, A.Wadas wrote:
> > Running Qmail on UltraWide disk and having at the moment no problems
> > makes me thinking
> > how to save money. Does it hurt significantly performance If I change
> > the disk to IDE one
> > Qmail is doing not more than 300 messages a day .
>
> You could do 300 messages a day off of a floppy disk.
>
> Chris
That is fine. So, does Qmail or any other mail server needs SCSI disks
at all. Is 10000 messages a day a good number or
still disk speed is not important?
Andrzej
Depends, are those 10,000 messages local deliveries? just queued mail
going out? I am using IDE drives for the qmail queue and NFS for
the maildir spool space. (On 3 PII/266-350's) and we deliver apx 140,000
messages a day. (Thats remote and local) ... most of those are incomming
local messages that go though the qmail queue and deliver onto the NFS
spool. Most of that is done on only one machine. The people reading all
share the same 3 machines, we have one other machine that is doing
listserv deliveries that I am not counting in my 140,000.
If I was delivering locally to local drives, I would use RAID to hold the
spool space, but probably still use ide to hold the queueing space,
nothing like hot swaping a drive and not losing any data or even having to
bring the spool space down. For the load I'm seeing on the IDE disks
currently doing queueing, I don't see any reason to move to SCSI.
If I was trying to deliver outbound millions of messages a day then I
would probably talk about moving to SCSI drives or maybe something faster
than that... but for general use why bother...
On the other hand speaking of delivering these huge mailing lists..
wouldn't it be more cost effective to purchas a second PC load linux and
use them both to split the queue load. :) When you can get an entire PC
for $500 and a SCSI card costs $120+SCSI disk... why not just purchase
another PC to split the load. That gains you 2 Ethernet channels/2 Memory
channels/2 xxx/redundance/failover and a boatload of other benifits...
Just my thoughts.
laters,
-d.
On Fri, 16 Apr 1999, A.Wadas wrote:
> Chris Johnson wrote:
>
> > On Fri, Apr 16, 1999 at 04:46:37PM +0000, A.Wadas wrote:
> > > Running Qmail on UltraWide disk and having at the moment no problems
> > > makes me thinking
> > > how to save money. Does it hurt significantly performance If I change
> > > the disk to IDE one
> > > Qmail is doing not more than 300 messages a day .
> >
> > You could do 300 messages a day off of a floppy disk.
> >
> > Chris
>
> That is fine. So, does Qmail or any other mail server needs SCSI disks
> at all. Is 10000 messages a day a good number or
> still disk speed is not important?
> Andrzej
>
>
+-----------------------------------------------------------------------+
| David Galbraith dgalb@ University Of New Mexico |
| Systems Analyst unm.edu (505)-277-8499 |
+-----------------------------------------------------------------------+
David A Galbraith CIRT wrote:
> Depends, are those 10,000 messages local deliveries? just queued mail
> going out? I am using IDE drives for the qmail queue and NFS for
> the maildir spool space. (On 3 PII/266-350's) and we deliver apx 140,000
> messages a day. (Thats remote and local) ... most of those are incomming
> local messages that go though the qmail queue and deliver onto the NFS
> spool. Most of that is done on only one machine. The people reading all
> share the same 3 machines, we have one other machine that is doing
> listserv deliveries that I am not counting in my 140,000.
>
> If I was delivering locally to local drives, I would use RAID to hold the
> spool space, but probably still use ide to hold the queueing space,
> nothing like hot swaping a drive and not losing any data or even having to
> bring the spool space down. For the load I'm seeing on the IDE disks
> currently doing queueing, I don't see any reason to move to SCSI.
>
> If I was trying to deliver outbound millions of messages a day then I
> would probably talk about moving to SCSI drives or maybe something faster
> than that... but for general use why bother...
>
> On the other hand speaking of delivering these huge mailing lists..
> wouldn't it be more cost effective to purchas a second PC load linux and
> use them both to split the queue load. :) When you can get an entire PC
> for $500 and a SCSI card costs $120+SCSI disk... why not just purchase
> another PC to split the load. That gains you 2 Ethernet channels/2 Memory
> channels/2 xxx/redundance/failover and a boatload of other benifits...
>
> Just my thoughts.
>
> laters,
> -d.
>
Thanks for answer. This was exactly what I was looking for.Andrzej
A.Wadas writes:
> That is fine. So, does Qmail or any other mail server needs SCSI disks
> at all. Is 10000 messages a day a good number or
> still disk speed is not important?
I always use SCSI disks, even where the speed is not an issue. I find that
on average, SCSI disks are more reliable than IDE[1], and I have fewer
problems with them.
After five years of trouble-free service, a few extra hundred bucks for a
SCSI disk do not seem to matter much.
--
Sam
[1]Provider to avoid certain troublesome brands.
>Hi. This is the qmail-send program at ns.crynwr.com.
>I tried to deliver a bounce message to this address, but the bounce bounced!
>
><[EMAIL PROTECTED]>:
>ezmlm-return: fatal: I do not accept messages at this address (#5.1.1)
>
>--- Below this line is the original bounce.
Russell, mind explaining?
ezmlm-return complains, presumably because it doesn't sit at the list
[EMAIL PROTECTED] I assume the list is [EMAIL PROTECTED] and the "Army"
has modified the envelope with the percent hack and thus _not_ bouncing
to the original envelope sender(?)
[EMAIL PROTECTED] works, [EMAIL PROTECTED] fails (but it appears
that pdam.crynwr.com = crynwr.com is a MX for ns.crynwr.com, but treats
ns.crynwr.com mail as local).
-Sincerely, Fred
(Frederik Lindberg, Infectious Diseases, WashU, St. Louis, MO, USA)
Fred Lindberg writes:
> >Hi. This is the qmail-send program at ns.crynwr.com.
> >I tried to deliver a bounce message to this address, but the bounce bounced!
> >
> ><[EMAIL PROTECTED]>:
> >ezmlm-return: fatal: I do not accept messages at this address (#5.1.1)
> >
> >--- Below this line is the original bounce.
>
> Russell, mind explaining?
>
> ezmlm-return complains, presumably because it doesn't sit at the list
> [EMAIL PROTECTED] I assume the list is [EMAIL PROTECTED] and the "Army"
> has modified the envelope with the percent hack and thus _not_ bouncing
> to the original envelope sender(?)
Yup. John Levine related in private mail that he sees that from time
to time also. Somebody's MTA autobogotifies the address with the %
hack when it sees that the MX doesn't match the hostname. I guess
they think it's helpful. Maybe ten years ago it would have been
helpful.
--
-russ nelson <[EMAIL PROTECTED]> http://crynwr.com/~nelson
Crynwr supports Open Source(tm) Software| PGPok | There is good evidence
521 Pleasant Valley Rd. | +1 315 268 1925 voice | that freedom is the
Potsdam, NY 13676-3213 | +1 315 268 9201 FAX | cause of world peace.
Hello List,
Is there a way to get qmail to deliver those messages which say (more or
less):
Your message could not be delivered to xxx. Do not send it again, your
message will remain in the mail queue and will be attempted to be
delivered for 7 more days.
Qmail obviously does that, but doesn´t kick back the message to sender
telling them what it is doing.
regards - eric
+ + + + + + + + + + + + + + + + + + + +
Spark Sistemas
- presentado por IWCC Argentina S.A.
Tel: 4702-1958
e-mail: [EMAIL PROTECTED]
+ + + + + + + + + + + + + + + + + + + +
I am still fighting the delivery problem - it keeps sayin that it
doesn't have permissions on the Maildir - I am still looking at it.
I am looking for anyones opinion on qmail and a larg load.
Our company has just been purchased by another larger ISP - they have
over 48,000 dial up users - all having a pop account.
How does qmail handle that kind of load on a round robin (multiple pop
boxes and smtp boxes)?
------------------------------------------------------------------------
----------------
James Beam
Support
PDQ.net ENTECH.com
http://www.pdq.net http://www.entech.com
------------------------------------------------------------------------
----------------
<<Jim Beam.vcf>>
Jim Beam.vcf
I am having trouble preventing Qmail from doing open relaying without
stopping mail service entirely. This is on a system that runs Listserv(R)
from the L-Soft Corporation. When I put the system's name (and aliases) in
a rcpthosts file, open relaying stops. The problem is that no one can send
e-mail to any account on this system. Qmail says that the sending system
is not allowed to relay. I do not understand this. In sendmail, I can stop
third party relaying without totally disabaling mail functionality on the
system, but I can't figure out how to do this with qmail. The Qmail FAQ
file was unclear on this subject, which is why I am asking here.
On Fri, Apr 16, 1999 at 11:12:35AM -0400, Stanley Horwitz wrote:
> I am having trouble preventing Qmail from doing open relaying without
> stopping mail service entirely. This is on a system that runs Listserv(R)
> from the L-Soft Corporation. When I put the system's name (and aliases) in
> a rcpthosts file, open relaying stops. The problem is that no one can send
> e-mail to any account on this system. Qmail says that the sending system
> is not allowed to relay. I do not understand this. In sendmail, I can stop
> third party relaying without totally disabaling mail functionality on the
> system, but I can't figure out how to do this with qmail. The Qmail FAQ
> file was unclear on this subject, which is why I am asking here.
What do you mean by "put the system's name (and aliases) in rcpthosts"?
rcpthosts should be a list of domains for which you're willing to receive mail
via SMTP. Some of these domains might be the same as your server's name and
aliases; then again, none of them may be, and your server may have aliases for
which you don't want to receive mail.
Make sure that your rcpthosts file contains only domains that are listed in
locals or virtualdomains, and domains for which you're acting as secondary mail
exchanger. Then you will be safe from unauthorized (or for that matter, any)
relaying, and you'll still be able to receive mail for any of the domains you
host.
Chris
At 01:42 PM Friday 4/16/99, Peter Samuel wrote:
>FYI
>
>I attempted to go to www.qmail.org but fat fingers sent me to
>www.qmail.com. This site is selling its mailing services under the
>ForeverMail banner.
And as for www.qmail.net...
whois shows that qmail.com was registered after the first beta of qmail but
before the 1.00 release.
Just another example of the free-for-all domain registration game I reckon.
Regards.
-----BEGIN PGP SIGNED MESSAGE-----
Hi folks,
I don't recall who asked about whether or not dashes were
permissed in qmail aliases. Well, I accidentally nuked the message before
I ran off to check.
Anyway, just wanted to pass along that I now have "rocket-scientist"
in use here as an alias to me and it works fine. Dunno what your problem
might be.
- -Jay
( ______
)) .-- "There's always time for a good cup of coffee." --. >===<--.
C|~~| (>-- Jay D. Dyson -- [EMAIL PROTECTED] --<) | = |-'
`--' `-- People who think NASA is fake view WWF as real. --' `-----'
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
iQCVAwUBNxdan82OVDpaKXD9AQHnbAP/V18JJQLsMYkHt5ZDGU/dT/m2PoM760fd
9hGdjl3e0LyA7gBlFGaGaODAdCmjKRJb57LDxgNuzQAyCcjHZ0X99mcvT0DAgbGs
kNBgssHKroCfAoJuov6oF/H9FOoNIBwQLF727lWOkHu9e0jF1qDgisWbXkPpV6A5
a9npSlZvxeU=
=b26f
-----END PGP SIGNATURE-----
jay,
as you can see in my address i have 'Marlon.Abao' which is aliased to my
username. However, when i place a 'Marlon-Abao: username' in the
'/etc/aliases' file and run newaliases, this entry just disappears :(
and so emails to this alias gets bounced back. how did you place your
entry in the '/etc/aliases' file?
-marlon
At 08:43 AM 4/16/99 -0700, Jay D. Dyson wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>
>Hi folks,
>
> I don't recall who asked about whether or not dashes were
>permissed in qmail aliases. Well, I accidentally nuked the message before
>I ran off to check.
>
> Anyway, just wanted to pass along that I now have "rocket-scientist"
>in use here as an alias to me and it works fine. Dunno what your problem
>might be.
>
>- -Jay
>
> ( ______
> )) .-- "There's always time for a good cup of coffee." --. >===<--.
> C|~~| (>-- Jay D. Dyson -- [EMAIL PROTECTED] --<) | = |-'
> `--' `-- People who think NASA is fake view WWF as real. --' `-----'
>
>-----BEGIN PGP SIGNATURE-----
>Version: 2.6.2
>
>iQCVAwUBNxdan82OVDpaKXD9AQHnbAP/V18JJQLsMYkHt5ZDGU/dT/m2PoM760fd
>9hGdjl3e0LyA7gBlFGaGaODAdCmjKRJb57LDxgNuzQAyCcjHZ0X99mcvT0DAgbGs
>kNBgssHKroCfAoJuov6oF/H9FOoNIBwQLF727lWOkHu9e0jF1qDgisWbXkPpV6A5
>a9npSlZvxeU=
>=b26f
>-----END PGP SIGNATURE-----
>
>
>
I bow to the masters of qmail...
My delivery problem was in my permissions.
Thanks for your patience.
------------------------------------------------------------------------
----------------
James Beam
Support
PDQ.net ENTECH.com
http://www.pdq.net http://www.entech.com
------------------------------------------------------------------------
----------------
<<Jim Beam.vcf>>
Jim Beam.vcf
The story:
----------
I've set up a mailserver RH Linux 5.1 running Qmail as my MTA.
The idea being to use OUTLOOK on the PC's to deliver mail to the LINUX
mailserver. The mailserver then dials-up my ISP and delivers all the
external e-mail and collects any new mail.
Now everything works with one exception.
When I send an external e-mail from OUTLOOK on one of the PC's to the LINUX
server via SMTP I get a delivery error saying that the user doesn't exist.
What should happen is that the external mail should be placed in the
~alias/pppdir/new and not bounced back.
However when I send internal mail from OUTLOOK it is delivered to the users
MAILDIR on the server fine.
If I send external email locally from the LINUX mailserver it is delivered
fine to the ~alias/pppdir/new waiting to be sent out to my ISP on the next
connection.
So please can some one shine some light as to why when I send external mail
from OUTLOOK via qmail-smtpd it is returned as undeliverable.
PS I start qmail-smtpd from inetd I know it isn't the preferred way of
doing it but we only have 10 terminal using the mailserver.
When I telnet to port 25 on the server it takes about 1.5 mins before
connecting. However if a ppp link is up it connects immediately. Any help
as to why that happens would also be nice.
Thanks.
Robert
On Fri, Apr 16, 1999 at 07:20:59PM +0100, Robert Harrison wrote:
> The story:
> ----------
> I've set up a mailserver RH Linux 5.1 running Qmail as my MTA.
>
> The idea being to use OUTLOOK on the PC's to deliver mail to the LINUX
> mailserver. The mailserver then dials-up my ISP and delivers all the
> external e-mail and collects any new mail.
>
>
> Now everything works with one exception.
>
> When I send an external e-mail from OUTLOOK on one of the PC's to the LINUX
> server via SMTP I get a delivery error saying that the user doesn't exist.
> What should happen is that the external mail should be placed in the
> ~alias/pppdir/new and not bounced back.
>
> However when I send internal mail from OUTLOOK it is delivered to the users
> MAILDIR on the server fine.
>
> If I send external email locally from the LINUX mailserver it is delivered
> fine to the ~alias/pppdir/new waiting to be sent out to my ISP on the next
> connection.
>
> So please can some one shine some light as to why when I send external mail
> from OUTLOOK via qmail-smtpd it is returned as undeliverable.
Does it really say that the user doesn't exist? Or does it say the domain isn't
in the list of rcpthosts? Your being able to inject mail locally but not via
qmail-smtpd almost certainly points to a relaying problem. Do you have
selective relaying configured, so that machines on your network can relay?
> PS I start qmail-smtpd from inetd I know it isn't the preferred way of
> doing it but we only have 10 terminal using the mailserver.
>
> When I telnet to port 25 on the server it takes about 1.5 mins before
> connecting. However if a ppp link is up it connects immediately. Any help
> as to why that happens would also be nice.
That's a reverse DNS lookup problem. You can either run local DNS to reverse
map your private IP addresses, or you can configure inetd not to do the reverse
mapping (and I don't know how to do that with inetd).
Chris
Chris,
Thanks for your pointer on the SMTP delay being caused by a time out on the
DNS I will reverse map to resolve.
I have just been into work to get a copy of the returned external mail and
the message is
Your message did not reach some or all of the intended recipients.
Subject: Testing External
Sent: 17/04/99
The following recipient(s) could not be reached:
'[EMAIL PROTECTED]' ....
This is what I get doing an qmail-smtp delivery from OUTLOOK to my LINUX
mailserver.
Somehow I want the external mail to be delivered to ~alias/pppdir is there
a .qmail config file I have to set up for sptp deliveries.
It all works fine when I send external mail from the mailserver. The
problem only occures with external mail addresses sent via SMTP to the
mailsever. These messages are being attempted to be delivered by qmail and
not put in ~alias/pppdir.
I'm probably doing something stupid, so kick me if I am.
Thanks Robert
----------
> From: Chris Johnson <[EMAIL PROTECTED]>
> To: Robert Harrison <[EMAIL PROTECTED]>
> Cc: [EMAIL PROTECTED]
> Subject: Re: SMTP Delivery Problem from Outlook to Qmail
> Date: 16 April 1999 20:48
>
> On Fri, Apr 16, 1999 at 07:20:59PM +0100, Robert Harrison wrote:
> > The story:
> > ----------
> > I've set up a mailserver RH Linux 5.1 running Qmail as my MTA.
> >
> > The idea being to use OUTLOOK on the PC's to deliver mail to the LINUX
> > mailserver. The mailserver then dials-up my ISP and delivers all the
> > external e-mail and collects any new mail.
> >
> >
> > Now everything works with one exception.
> >
> > When I send an external e-mail from OUTLOOK on one of the PC's to the
LINUX
> > server via SMTP I get a delivery error saying that the user doesn't
exist.
> > What should happen is that the external mail should be placed in the
> > ~alias/pppdir/new and not bounced back.
> >
> > However when I send internal mail from OUTLOOK it is delivered to the
users
> > MAILDIR on the server fine.
> >
> > If I send external email locally from the LINUX mailserver it is
delivered
> > fine to the ~alias/pppdir/new waiting to be sent out to my ISP on the
next
> > connection.
> >
> > So please can some one shine some light as to why when I send external
mail
> > from OUTLOOK via qmail-smtpd it is returned as undeliverable.
>
> Does it really say that the user doesn't exist? Or does it say the domain
isn't
> in the list of rcpthosts? Your being able to inject mail locally but not
via
> qmail-smtpd almost certainly points to a relaying problem. Do you have
> selective relaying configured, so that machines on your network can
relay?
>
> > PS I start qmail-smtpd from inetd I know it isn't the preferred way of
> > doing it but we only have 10 terminal using the mailserver.
> >
> > When I telnet to port 25 on the server it takes about 1.5 mins before
> > connecting. However if a ppp link is up it connects immediately. Any
help
> > as to why that happens would also be nice.
>
> That's a reverse DNS lookup problem. You can either run local DNS to
reverse
> map your private IP addresses, or you can configure inetd not to do the
reverse
> mapping (and I don't know how to do that with inetd).
>
> Chris
Is there anyway to forward an smtp request from one machine to another?
If my customer's have their pop clients configured with the smtp server
"mail.mydomain.com" can the machine "mail.mydomain.com" forward smtp
requests to "mail2.mydomain.com" ?
I'd like to move outgoing mail services to another machine without
affecting the incoming mail. Because of my previous setup all of my
clients have both POP and SMTP servers set to "mail.mydomain.com"
Any ideas?
Rob Genovesi
[EMAIL PROTECTED]
Rob Genovesi wrote:
> I'd like to move outgoing mail services to another machine without
> affecting the incoming mail. Because of my previous setup all of my
> clients have both POP and SMTP servers set to "mail.mydomain.com"
>
>
If you can discriminate by incoming IP address you can direct traffic to a
different
machine/port combination (all customers come in over local dial-up vs.
incoming email
from net coming in from a different address range). Even so you would have
to be able
to support both in & out on both machines for those who didn't fit the
above model.
It's hard to think of all eventualities when you build services, but since
A records and
CNAMEs are free you might as well dream up as many as you want.
(You could achieve a similar effect as above by having split-horizon dns,
might be
easier too).
This took me a little while to get right. I use listserv 1.8d and qmail.
rcpthosts has all my usual aliases: localhost, listserv.domain.name,
eachmachine.domain.name, etc.
Then for relaying, I have my entire local subnet in /etc/hosts.allow:
tcp-env: my.sub.net. : setenv RELAYCLIENT
and smtpd in inetd.conf:
smtp stream tcp nowait qmaild /path/to/tcp-env tcp-env
/path/to/qmail-smtpd
The trick is to make ABSOLUTELY sure that -DPROCESS_OPTIONS is enabled for
tcp wrappers. Even if you think it is, double-check it.
Scott Swanson
Sysadmin, CTW Online
[EMAIL PROTECTED]
http://www.ctw.org/
Hi.
I'm curious, why do you guys use qmail with ListServ, and not LSMTP?
Specifically, did you find qmail to be faster?
David.
At 04:16 PM 4/16/99 -0400, Swanson, Scott wrote:
>This took me a little while to get right. I use listserv 1.8d and qmail.
>
>rcpthosts has all my usual aliases: localhost, listserv.domain.name,
>eachmachine.domain.name, etc.
>
>Then for relaying, I have my entire local subnet in /etc/hosts.allow:
>
>tcp-env: my.sub.net. : setenv RELAYCLIENT
>
>and smtpd in inetd.conf:
>
>smtp stream tcp nowait qmaild /path/to/tcp-env tcp-env
>/path/to/qmail-smtpd
>
>The trick is to make ABSOLUTELY sure that -DPROCESS_OPTIONS is enabled for
>tcp wrappers. Even if you think it is, double-check it.
>
>Scott Swanson
>Sysadmin, CTW Online
>[EMAIL PROTECTED]
>http://www.ctw.org/
>
>
>
Hi,
I was digging around on the qmail web-page when I came across the
announcement of Bruce Guenter's mini-qmail implementation "nullmailer"
which is described as having "...a queue for more reliablity".
What is meant by "more reliable" in this example? Is the standard
mini-qmail setup "unreliable"?
Our incoming mail uses the following setup:
incoming +---------+ +-----------+
smtp | bastion | qmqp | internal |
--------> | host | ------> | mail host |
+---------+ +-----------+
mini-qmail full qmail
( mail received ( mail received
by qmail-smtpd ) by qmail-qmqpd )
>From reading the qmqpd documentation, this process should be fail-safe,
ie the remote host making the SMTP connection only gets the "all clear"
when the mail has been successfully delivered into the remote queue.
Am I missing something, or is there a potential blackhole somewhere in
this setup that a queue on the bastion host would prevent?
R.
PS. I *think* I've got my news2mail gateway working properly now, ie
the header should now be correct and not contain any "internal"
addresses. Please let me know if you think something is still wrong.
One problem I still have is filtering out Cancel messages. Currently,
if I post to my internal newsgroup and then cancel the message, both the
original message and the cancel message are sent back to the list.
Simple workaround: don't cancel any messages! However, I'd prefer
something a little more fool proof. Is there any program that looks in
the headers for a specified field, ie a bit like "iftocc" but for a
specific header? In this case, I could look out for either:
Subject: cancel ...
or
Control: cancel ...
It would probably not be a bad idea to consign all control messages to
the bit bucket; any suggestions as to how?
--
Two rules to success in life:
1. Don't tell people everything you know.
-- Sassan Tat
At 11:12 PM Friday 4/16/99, Robin Bowes wrote:
>Hi,
>
>I was digging around on the qmail web-page when I came across the
>announcement of Bruce Guenter's mini-qmail implementation "nullmailer"
>which is described as having "...a queue for more reliablity".
>
>What is meant by "more reliable" in this example? Is the standard
>mini-qmail setup "unreliable"?
It's not relevant to your setup. There is some question as to whether a
local queue is more reliable than a remote queue accessible via qmqpc.
Who is right all depends on how well the local/remote systems are maintained.
>
>Our incoming mail uses the following setup:
>
> incoming +---------+ +-----------+
> smtp | bastion | qmqp | internal |
> --------> | host | ------> | mail host |
> +---------+ +-----------+
> mini-qmail full qmail
> ( mail received ( mail received
> by qmail-smtpd ) by qmail-qmqpd )
>
>>From reading the qmqpd documentation, this process should be fail-safe,
That's a perfectly legitimate setup. The remote sending site only gets a
"250 ok" if the internal mail host queues the mail successfully.
Regards.
Mark Delany wrote:
>
> At 11:12 PM Friday 4/16/99, Robin Bowes wrote:
> >Hi,
> >
> >I was digging around on the qmail web-page when I came across the
> >announcement of Bruce Guenter's mini-qmail implementation "nullmailer"
> >which is described as having "...a queue for more reliablity".
> >
> >What is meant by "more reliable" in this example? Is the standard
> >mini-qmail setup "unreliable"?
>
> It's not relevant to your setup. There is some question as to whether a
> local queue is more reliable than a remote queue accessible via qmqpc.
Can you give an example of when it might be relevant?
>
> Who is right all depends on how well the local/remote systems are maintained.
>
> >
> >Our incoming mail uses the following setup:
> >
> > incoming +---------+ +-----------+
> > smtp | bastion | qmqp | internal |
> > --------> | host | ------> | mail host |
> > +---------+ +-----------+
> > mini-qmail full qmail
> > ( mail received ( mail received
> > by qmail-smtpd ) by qmail-qmqpd )
> >
> >>From reading the qmqpd documentation, this process should be fail-safe,
>
> That's a perfectly legitimate setup. The remote sending site only gets a
> "250 ok" if the internal mail host queues the mail successfully.
That was my feeling on reading the docs.
So, what's this "nullmailer" all about then? Presumably it applies to a
different setup than mine? Bruce?
R.
--
Two rules to success in life:
1. Don't tell people everything you know.
-- Sassan Tat
>> It's not relevant to your setup. There is some question as to whether a
>> local queue is more reliable than a remote queue accessible via qmqpc.
>
>Can you give an example of when it might be relevant?
Mainly command-line based UA's that don't check the exit code of the mail
submission program.
>So, what's this "nullmailer" all about then? Presumably it applies to a
>different setup than mine? Bruce?
I guess. There was some talk about some sort of half-way house between qmqpc
and qmail. Presumably queueing only on failure of qmqpc or somesuch and the
queue presumably being substantially simpler than the real one.
Regards.
Mark Delany ([EMAIL PROTECTED]) wrote:
....
: In practice I'm told that some select() implementations are buggy in this
: regard and may simply return "writable" if a single byte will fit. I have no
: direct evidence of this though.
I don't think this is the reason for the single byte, because
multiple bytes -are- being passed through the pipe, though only
the first is used for the job index (del_dochan()). At any rate,
a block on a write will affect only qmail-?spawn, not qmail-send,
and qmail-send will eventually unblock the pipe by reading from it.
Should not be a problem making it 2 bytes instead. An alternative
might be, if you don't feel like modifying both qmail-?spawns as
well as qmail-send, to modify only qmail-send to run 2 (or 3 or
more) qmail-rspawns in addition to the one lspawn.
-harold
It wasn't so much the blocking I was thinking of (and I confess ignorance on
this front) it was on the basis that multiple processes are writing to the
same pipe, thus a 2 byte write could get interspersed with another two byte
write by another process.
But as I say. I have not actually studied the details of the pipe arrangements
between -send and -[lr]spawn, so this may be a complete red herring.
Regards.
At 10:46 AM Saturday 4/17/99, wrote:
>Mark Delany ([EMAIL PROTECTED]) wrote:
>....
>
>: In practice I'm told that some select() implementations are buggy in this
>: regard and may simply return "writable" if a single byte will fit. I
have no
>: direct evidence of this though.
>
>I don't think this is the reason for the single byte, because
>multiple bytes -are- being passed through the pipe, though only
>the first is used for the job index (del_dochan()). At any rate,
>a block on a write will affect only qmail-?spawn, not qmail-send,
>and qmail-send will eventually unblock the pipe by reading from it.
>
>Should not be a problem making it 2 bytes instead. An alternative
>might be, if you don't feel like modifying both qmail-?spawns as
>well as qmail-send, to modify only qmail-send to run 2 (or 3 or
>more) qmail-rspawns in addition to the one lspawn.
>
>-harold
>
On Fri, 16 Apr 1999, Mark Delany wrote:
> It wasn't so much the blocking I was thinking of (and I confess ignorance on
> this front) it was on the basis that multiple processes are writing to the
> same pipe, thus a 2 byte write could get interspersed with another two byte
> write by another process.
>
> But as I say. I have not actually studied the details of the pipe arrangements
> between -send and -[lr]spawn, so this may be a complete red herring.
As long as you keep your writes small enough on a POSIX-compliant box,
they're guaranteed to be atomic, and not split up:
==============================================================================
6.2.4 Atomic Operations with Pipes
In order for an operation to be considered ``atomic'', it must not be
interrupted for any reason at all. The entire operation occurs at once.
The POSIX standard dictates in /usr/include/posix1_lim.h that the maximum
buffer size for an atomic operation on a pipe is:
#define _POSIX_PIPE_BUF 512
Up to 512 bytes can be written or retrieved from a pipe atomically.
Anything that crosses this threshold will be split, and not atomic. Under
Linux, however, the atomic operational limit is defined in
``linux/limits.h'' as:
#define PIPE_BUF 4096
As you can see, Linux accommodates the minimum number of bytes required by
POSIX, quite considerably I might add. The atomicity of a pipe operation
becomes important when more than one process is involved (FIFOS). For
example, if the number of bytes written to a pipe exceeds the atomic limit
for a single operation, and multiple processes are writing to the pipe,
the data will be ``interleaved'' or ``chunked''. In other words, one
process may insert data into the pipeline between the writes of another.
==============================================================================
I could not determine the semantics of select(), on this topic.
At 11:59 PM Friday 4/16/99, Sam wrote:
>On Fri, 16 Apr 1999, Mark Delany wrote:
>
>> It wasn't so much the blocking I was thinking of (and I confess
ignorance on
>> this front) it was on the basis that multiple processes are writing to the
>> same pipe, thus a 2 byte write could get interspersed with another two byte
>> write by another process.
>>
>> But as I say. I have not actually studied the details of the pipe
>arrangements
>> between -send and -[lr]spawn, so this may be a complete red herring.
>
>As long as you keep your writes small enough on a POSIX-compliant box,
>they're guaranteed to be atomic, and not split up:
Erum. But that's what my original response was all about. That conformance
to this is not universal.
Please read the original response or alternatively have a look at:
ftp://koobera.math.uic.edu/www/docs/unixport.html
Regards.
>
>==============================================================================
>6.2.4 Atomic Operations with Pipes
>
>In order for an operation to be considered ``atomic'', it must not be
>interrupted for any reason at all. The entire operation occurs at once.
>The POSIX standard dictates in /usr/include/posix1_lim.h that the maximum
>buffer size for an atomic operation on a pipe is:
>
> #define _POSIX_PIPE_BUF 512
>
>Up to 512 bytes can be written or retrieved from a pipe atomically.
>Anything that crosses this threshold will be split, and not atomic. Under
>Linux, however, the atomic operational limit is defined in
>``linux/limits.h'' as:
>
> #define PIPE_BUF 4096
>
>As you can see, Linux accommodates the minimum number of bytes required by
>POSIX, quite considerably I might add. The atomicity of a pipe operation
>becomes important when more than one process is involved (FIFOS). For
>example, if the number of bytes written to a pipe exceeds the atomic limit
>for a single operation, and multiple processes are writing to the pipe,
>the data will be ``interleaved'' or ``chunked''. In other words, one
>process may insert data into the pipeline between the writes of another.
>==============================================================================
>
>I could not determine the semantics of select(), on this topic.
>
>
Hello,
I notice there extra field delimiters in the users/assign file. Are these
used for anything? If I start putting data in them will it mess anything
else up? Can I add fields?
If anyone has tried this I would like to hear of your experiences.
Thanks,
Ted
On Fri, Apr 16, 1999 at 09:55:57PM -0600, Theodore Cekan wrote:
> Hello,
>
> I notice there extra field delimiters in the users/assign file. Are these
> used for anything? If I start putting data in them will it mess anything
> else up? Can I add fields?
There aren't any extras. This is from the qmail-users man page:
A simple assignment is a line of the form
=local:user:uid:gid:homedir:dash:ext:
Here local is an address; user, uid, and gid are the
account name, uid, and gid of the user in charge of local;
and messages to local will be controlled by home-
dir/.qmaildashext.
Chris
Our company has had MS Exchange in place for a couple of years now and the corporate
users
are pretty enamored of it's scheduling capabilites. In the past we have not had an
Internet connection until now. I have a Linux box connected to the Internet and have
started setting up qmail as our 'connected' mail server.
I would like to have inbound mail passed on to the Exchange server that's inhouse here
(on a 10.x.x.x reserved IP) as well as have the outbound mail delivered to qmail to
deal
with from there...
Anyone out there done this before? If so, if you have any quick tips to get me started
in the right direction that'd be great.
Thanks if so >-)
Scott Burkhalter
