qmail Digest 6 Aug 1999 10:00:00 -0000 Issue 720
Topics (messages 28596 through 28649):
Virtual domains
28596 by: Bernat Ginard <[EMAIL PROTECTED]>
28610 by: Ken Jones <[EMAIL PROTECTED]>
Problem with installation.
28597 by: "Mark Weinem" <[EMAIL PROTECTED]>
Cc: header rewriting.
28598 by: Kamil Andrusz <[EMAIL PROTECTED]>
28614 by: "Fred Lindberg" <[EMAIL PROTECTED]>
qmail & DNS
28599 by: Simon Rae <[EMAIL PROTECTED]>
28600 by: "Petr Novotny" <[EMAIL PROTECTED]>
qmail-newu: fatal: bad format in users/assign
28601 by: [EMAIL PROTECTED]
Backward Alias System
28602 by: Magnus Bodin <[EMAIL PROTECTED]>
28606 by: Russell Nelson <[EMAIL PROTECTED]>
spanning lines in .qmail
28603 by: Russell Nelson <[EMAIL PROTECTED]>
non-existant host - defered?!
28604 by: "David Dyer-Bennet" <[EMAIL PROTECTED]>
28608 by: "Petr Novotny" <[EMAIL PROTECTED]>
28609 by: "David Dyer-Bennet" <[EMAIL PROTECTED]>
28612 by: "Petr Novotny" <[EMAIL PROTECTED]>
Fw: spanning lines in .qmail
28605 by: "Leon Vismer" <[EMAIL PROTECTED]>
28607 by: "Chris Garrigues" <[EMAIL PROTECTED]>
28613 by: Dave Kitabjian <[EMAIL PROTECTED]>
28615 by: Brad Shelton <[EMAIL PROTECTED]>
28618 by: Dave Kitabjian <[EMAIL PROTECTED]>
28619 by: [EMAIL PROTECTED] ()
can 'alias' run programs?
28611 by: Brian Reichert <[EMAIL PROTECTED]>
New qmail-ldap patch release
28616 by: Andre Oppermann <[EMAIL PROTECTED]>
M$ Exchange -> qmail
28617 by: "Robertson, William" <[EMAIL PROTECTED]>
SMTP/X400 conversion failure
28620 by: "Tim Hunter" <[EMAIL PROTECTED]>
28636 by: <[EMAIL PROTECTED]>
how to setup qmail in a chroot environment?
28621 by: J�rg St�dele <[EMAIL PROTECTED]>
28635 by: Peter van Dijk <[EMAIL PROTECTED]>
qmail+jbuce patch--DNS messed up?
28622 by: Peter Green <[EMAIL PROTECTED]>
28625 by: "Aaron L. Meehan" <[EMAIL PROTECTED]>
28626 by: Peter Green <[EMAIL PROTECTED]>
28629 by: "Adam D . McKenna" <[EMAIL PROTECTED]>
28631 by: "Aaron L. Meehan" <[EMAIL PROTECTED]>
28632 by: Peter Green <[EMAIL PROTECTED]>
mail volume
28623 by: Troy Morrison <[EMAIL PROTECTED]>
smtp problem with quotes on hostname end of hostname
28624 by: [EMAIL PROTECTED]
Qmail throughput
28627 by: Jim Arnott <[EMAIL PROTECTED]>
28630 by: Daemeon Reiydelle <[EMAIL PROTECTED]>
28633 by: Mark Delany <[EMAIL PROTECTED]>
28634 by: Mark Delany <[EMAIL PROTECTED]>
28637 by: Jim Arnott <[EMAIL PROTECTED]>
28638 by: Jim Arnott <[EMAIL PROTECTED]>
28639 by: Mark Delany <[EMAIL PROTECTED]>
28640 by: "David Dyer-Bennet" <[EMAIL PROTECTED]>
28641 by: "Richard Shetron" <[EMAIL PROTECTED]>
28642 by: Daemeon Reiydelle <[EMAIL PROTECTED]>
28643 by: [EMAIL PROTECTED]
28644 by: [EMAIL PROTECTED] ()
28648 by: "Petr Novotny" <[EMAIL PROTECTED]>
using sendpage with qmail
28628 by: Brian Reichert <[EMAIL PROTECTED]>
Hi again!
28645 by: "Marc-Adrian Napoli" <[EMAIL PROTECTED]>
28646 by: M Lyons <[EMAIL PROTECTED]>
28649 by: "Petr Novotny" <[EMAIL PROTECTED]>
HELP: Distributed mail system
28647 by: Jason Brown <[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]
----------------------------------------------------------------------
Hi all,
I'm trying to configure qmail to use virtual
domain. I use file control/virtualdomains and all
works fine.
The problem is we want that qmail substitute
a Windows NT mail server were clients access the server
via pop3. With Windows NT the server has a IP for each
domain and the pop3 server identifies the domain looking
at which one of its address has connected the client and
there isn't the need of use account names of the style
domain-user (almost for user point of view), is this
possible with qmail. It's becouse it would represent
a big work to instruct the users to reconfigure their
mail reader: most of them are unable to change it
without on line help.
--
Bernat Ginard Llad�
mailto:[EMAIL PROTECTED] http://www.kaos.es
Bernat Ginard wrote:
>
> Hi all,
>
> I'm trying to configure qmail to use virtual
> domain. I use file control/virtualdomains and all
> works fine.
>
> The problem is we want that qmail substitute
> a Windows NT mail server were clients access the server
> via pop3. With Windows NT the server has a IP for each
> domain and the pop3 server identifies the domain looking
> at which one of its address has connected the client and
> there isn't the need of use account names of the style
> domain-user (almost for user point of view), is this
> possible with qmail. It's becouse it would represent
> a big work to instruct the users to reconfigure their
> mail reader: most of them are unable to change it
> without on line help.
>
The new developement version of vchkpw ( 3.4.6 ) handles this.
The old way was the user would change thier authentication
pop name to be user%domain. (That still works too).
The new way follows this logic:
1. If the pop authentication name does not contain a "%domainname" then
a) An IP to domain name lookup is done on the IP address the user
connected to on the pop server.
b) A check is made to determine if the domainname is in the
~control/locals file
If it is in the locals file then
The user is authenticated against:
1) /etc/passwd if that fails then
2) the pop only users file
c) If it is not in the locals file then
A check is made to see if the virtual domain exists
If it exists the user is authenticated against the virtual
domains pop only users file.
2> If the pop authentication name contains a "%domainname" then (this is
the old way)
A check is made to see if a virtual domain with the name exists
If it exists the user is authenticated against the virtual
domains pop only users file.
We have tested it locally and will start testing it on other machines in
the next
week. So far everything works great.
Ken Jones
http://www.inter7.com/vchkpw/
Hi,
> There seems to be some problem with my DNS.
Not really ;-)
Do (as root):
# echo "yourhostname.yourdomain.your-top-level-domain" > /var/qmail/control/me
My full hostname is pandora.plagegeister.de , i did:
# echo "pandora.plagegeister.de" > /var/qmail/control/me
regards,
Mark
Hi there =)
In my company we are trying to achieve header rewriting of outgoing mails,
along with mail routing. The situation is as follows:
qmail.mail.serv
/ \
/ \
/ \
/ > Internet
Intranet
We need to rewrite the headers for all mail going to the Internet to
Internet e-mail addresses. That means that we have to rewrite To:, From: and
also Cc:. For Intranet we rewrite to Intranet e-mail addresses (some of our
users use Intranet adresses some use Internet... total mess =).
I've created a script which does all necessary rewriting.
The mail is parsed throught the script using the "fixup" virtual domain.
Headers are rewritten using reformail.
Every thing goes ok until qmail rewrites the Cc: header, then it sends
another copy of the mail to all the recipients. This means that if I have
three Cc's and one To: then each user will get four mails.
It's not the fault of the script. It was debugged a million times by now
and we didn't find anything wrong in it. When we remove the part responsible
for Cc: rewriting everything works fine, but we need to rewrite the Cc's.
Does anyone have a slightest idea what I am doing wrong ?
Sorry if that was chaotic =)
GreetZ WiZzArD
--
[ Kamil Andrusz : nick WiZzArD : Admin : Lido Technology : A Lufthansa Company]
[ [EMAIL PROTECTED] : [EMAIL PROTECTED] : Ngoah! Shwurzbung ]
[ God. root. what's the difference. ]
On Thu, 5 Aug 1999 13:39:10 +0200, Kamil Andrusz wrote:
> Every thing goes ok until qmail rewrites the Cc: header, then it sends
>another copy of the mail to all the recipients. This means that if I have
>three Cc's and one To: then each user will get four mails.
If I understand correctly, you reinject the message with a command that
parses the addresses from the message itself. Use qmail-inject instead
of the sendmail interface and use the appropriate qmail-inject option,
placing the recipient address on the command line. If I'm wrong, it
might help if you post the rewriting script.
>From the qmail-inject man page:
OPTIONS
-a Send the message to all addresses given as recip
arguments; do not use header recipient addresses.
-h Send the message to all header recipient addresses.
For non-forwarded messages, this means the
addresses listed under To, Cc, Bcc, Apparently-To.
For forwarded messages, this means the addresses
listed under Resent-To, Resent-Cc, Resent-Bcc. Do
not use any recip arguments.
-A (Default.) Send the message to all addresses given
as recip arguments. If no recip arguments are sup-
plied, send the message to all header recipient
addresses.
-H Send the message to all header recipient addresses,
and to all addresses given as recip arguments.
-Sincerely, Fred
(Frederik Lindberg, Infectious Diseases, WashU, St. Louis, MO, USA)
Just a quickie.
Does qmail use resolv.conf to to its DNS lookups? If not, then what's
the process?
Simon Rae
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
> Does qmail use resolv.conf to to its DNS lookups? If not, then what's the
> process?
It does - at least it has to know what the nameserver's IP is! (It
doesn't use /etc/hosts etc. but that's a different fairy-tale.)
-----BEGIN PGP SIGNATURE-----
Version: PGP 6.0.2 -- QDPGP 2.60
Comment: http://community.wow.net/grt/qdpgp.html
iQA/AwUBN6mI7FMwP8g7qbw/EQKm+gCg5hjvcx5exDnZK9DbOHmvxHjK784An1Q7
VawZbjUpcpaQuAGul8oRD5Cw
=lHzj
-----END PGP SIGNATURE-----
--
Petr Novotny, ANTEK CS
[EMAIL PROTECTED]
http://www.antek.cz
PGP key ID: 0x3BA9BC3F
-- Don't you know there ain't no devil there's just God when he's drunk.
[Tom Waits]
On Thu, 5 Aug 1999, Nguyen Dang Phuoc Dong wrote:
[snip]
> Then I run qmail-newu and it says "qmail-newu: fatal: bad format in
> users/assign".
This is almost certainly the missing '.' line required at the end. I
prefer to create the users/assign file automatically, using qmail-pw2u.
You can put your extra records for users/assign in users/append and never
have to worry about the dot line.
> I had read the qmail-users, qmail-newu manual page carefully. So, I don't
> know why it doesn't work.
>
> Please tell me what's wrong in my configuration. Thank you very much.
>
> Best regards,
>
> Dong
>
> ----------------------------------------------------------------
> Nguyen Dang Phuoc Dong
> Phuong Nam Net. - System Administrator.
>
>
>
--
"Life is much too important to be taken seriously."
Thomas Erskine <[EMAIL PROTECTED]> (613) 998-2836
On Thu, 5 Aug 1999, Daniel Callan wrote:
> >> I know that the archive is full of alias problems but so far I cannot
> >> seem to find someone mentioning the simplest (and worst) problem of
> >> them all: If an account exists that is the same name as a virtual-domain
> >> specific alias (or even any alias for that matter), the account takes
> >> preference over the alias and the mail goes straight to the account.
> >
> >This is not correct.
>
> Which part?
local is never virtual. (See other answers in this thread)
> All of the virtual domains I'm using are in control/locals AND
> in control/virtdomains. If they weren't, I couldn't use them as domains
which is your problem.
try to run qmail-lint which you can find on www.qmail.org
> for all the local accounts. None of the domains are soley for either
> purpose, they are all used for BOTH accounts and aliases.
>
> Point is, Sendmail was able to have virtual domains apply to all
> known accounts and aliases (WITH ALIASES CHECKED BEFORE ACCOUNTS).
> >
> >> Am I the only who thinks this is completey backwards???
> >
> >Unfortunately not.
>
> Glad to hear that at least ;-)
I meant that there are too many users that are confused like this about
local and virtual domain handling. (No offense intended).
If you describe EXACTLY what you want, it will be much easier to meet
the help demands. Take 2-3 domains together if you really want to merge
these configurations together. I usually don't.
/magnus
Daniel Callan writes:
> Hi all,
>
> I know that the archive is full of alias problems but so far I cannot
> seem to find someone mentioning the simplest (and worst) problem of
> them all: If an account exists that is the same name as a virtual-domain
> specific alias (or even any alias for that matter), the account takes
> preference over the alias and the mail goes straight to the account.
>
> Am I the only who thinks this is completely backwards???
It's not backwards. It's not forwards either. It's an artifact of
the way qmail resolves email addresses. First it determines the user,
and the user delivers/forwards the mail. Qmail has a separate user
for aliases. So qmail has to decide whether the mail goes to a user
or an alias before it knows the complete list of aliases. It can't do
this in advance of actually delivering the mail because of -default
wildcarding.
BTW, there's no way to implement (the RFC821 commands) VRFY or EXPN if
the matching user has a matching \.qmail.*default file. OTOH, it
would be quite possible to have VRFY or EXPN consult a cdb for the
right response to return. That is, if you *really* want to tell
anyone who walks up what your valid email addresses are.
> We have got all the fastforward/dot-forward packages included and have
> still been suffering from this for almost a year now.
Oh, well you can't have been suffering all that much, otherwise you
would have paid me to relieve your suffering. ( sorry, it's an old
economist's joke. Two economists are walking past a car dealership.
One economist says to the other "I'd give anything for that Porsche."
The other economist says "Obviously not.")
> However, we might have (for example) 5
> different aliases for "design"
> ie: [EMAIL PROTECTED] --> account1
> [EMAIL PROTECTED] --> account2
> [EMAIL PROTECTED] --> account3
> [EMAIL PROTECTED] --> account4
> [EMAIL PROTECTED] --> account5
>
> Now, it only takes one of them to unknowingly create a "design"
> account (thru the adduser system) to completely stuff things up and
> give me 5 complaining customers in one hit. Mail for "design@anything"
> will ONLY go to the mailbox (completely ignoring the aliases).
Sounds like you aren't using virtualdomains. Do this:
cd /var/qmail
cat >>control/virtualdomains <<EOF
dom1.com:alias-dom1
dom2.com:alias-dom2
dom3.com:alias-dom3
dom4.com:alias-dom4
dom5.com:alias-dom5
EOF
echo '&account1' >~alias/.qmail-dom1-design
echo '&account2' >~alias/.qmail-dom2-design
echo '&account3' >~alias/.qmail-dom3-design
echo '&account4' >~alias/.qmail-dom4-design
echo '&account5' >~alias/.qmail-dom5-design
kill -HUP `pidof qmail-send`
--
-russ nelson <[EMAIL PROTECTED]> http://crynwr.com/~nelson
Crynwr sells support for free 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!
Adam D . McKenna writes:
> | echo "Dear $SENDER,
> Thank You for joining our e-newsletter. This e-mail is a confirmation that
> you have been added to our e-newsletter mailing list.
> "
> Is there a way to span lines like this?
No. Either put it into a script, or put \n's into your echo.
Personally, I think that a leading space following a program delivery
ought to be appended to the command before it's handed to /bin/sh, but
that's just me.
--
-russ nelson <[EMAIL PROTECTED]> http://crynwr.com/~nelson
Crynwr sells support for free 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!
Ira Abramov <[EMAIL PROTECTED]> writes on 5 August 1999 at 08:49:54 +0300
>
>
> I just noticed messages to non-existant hosts (due to typos or whatever)
> don't get bounced, but instead stay stuck in the queue for the full
> lifetime (1 week here). any REALLY good reaon for that?
A host that looks non-existent could have have its DNS missing or
down, or its link down. Keeping it for a week gives them a chance to
get it back up.
On the other hand, of course it means you don't discover you mistyped
the address for a week.
I think a week queue lifetime is simply too long, myself. Haven't
changed it here yet, though.
--
David Dyer-Bennet ***NOTE ADDRESS CHANGES*** [EMAIL PROTECTED]
http://dd-b.lighthunters.net/ (photos) Minicon: http://www.mnstf.org/minicon
http://www.dd-b.net/dd-b (sf) http://ouroboros.demesne.com/ Ouroboros Bookworms
Join the 20th century before it's too late!
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
> A host that looks non-existent could have have its DNS missing or
> down, or its link down. Keeping it for a week gives them a chance to get
> it back up.
>
> On the other hand, of course it means you don't discover you mistyped the
> address for a week.
That's plainly not true. Only if you - by chance - hit an address for
which you can't get an authoritative "don't exist" answer, the mail
gets deferred.
Just now I tested to mail (positively non-existing)
[EMAIL PROTECTED] and the mail was back bounced in a second.
-----BEGIN PGP SIGNATURE-----
Version: PGP 6.0.2 -- QDPGP 2.60
Comment: http://community.wow.net/grt/qdpgp.html
iQA/AwUBN6m9qVMwP8g7qbw/EQJJ/ACgrQGZy53SHSNgTnkoxLILrd8fMysAoNTM
vQT1lWb6BRIVdaynlkVJGe9v
=EPkG
-----END PGP SIGNATURE-----
--
Petr Novotny, ANTEK CS
[EMAIL PROTECTED]
http://www.antek.cz
PGP key ID: 0x3BA9BC3F
-- Don't you know there ain't no devil there's just God when he's drunk.
[Tom Waits]
Petr Novotny <[EMAIL PROTECTED]> writes on 5 August 1999 at 17:36:58 +0100
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> > A host that looks non-existent could have have its DNS missing or
> > down, or its link down. Keeping it for a week gives them a chance to get
> > it back up.
> >
> > On the other hand, of course it means you don't discover you mistyped the
> > address for a week.
>
> That's plainly not true. Only if you - by chance - hit an address for
> which you can't get an authoritative "don't exist" answer, the mail
> gets deferred.
>
> Just now I tested to mail (positively non-existing)
> [EMAIL PROTECTED] and the mail was back bounced in a second.
The cases I described would mostly be soft failures, and would wait a
week. If you can get an authoritative DNS answer and the answer is
NO, that's a hard failure and it will bounce; but if the main DNS is
down or connectivity to it is down, you get a non-authoritative NO,
and qmail waits for better.
--
David Dyer-Bennet ***NOTE ADDRESS CHANGES*** [EMAIL PROTECTED]
http://dd-b.lighthunters.net/ (photos) Minicon: http://www.mnstf.org/minicon
http://www.dd-b.net/dd-b (sf) http://ouroboros.demesne.com/ Ouroboros Bookworms
Join the 20th century before it's too late!
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
> The cases I described would mostly be soft failures, and would wait a
> week. If you can get an authoritative DNS answer and the answer is NO,
> that's a hard failure and it will bounce; but if the main DNS is down or
> connectivity to it is down, you get a non-authoritative NO, and qmail
> waits for better.
Yep. That was my point from the very beginning. This behaviour is
correct as far as I am concerned.
I thought you tried to argue that you misspell an address and will
not learn before a week - but that's not too probable that a
misspelled address produces temporary DNS error.
(OTOH, now my queue contains quite a few mails to a host with a
lame server - top-level servers for .it know the domain but the
designated servers never heard of it. This is a soft error too.)
-----BEGIN PGP SIGNATURE-----
Version: PGP 6.0.2 -- QDPGP 2.60
Comment: http://community.wow.net/grt/qdpgp.html
iQA/AwUBN6nFQlMwP8g7qbw/EQJGoACeKUCMJfdyIi7O9Jo+9afGYL2GzZAAmgME
ial4S+sIv+VL6Eyz2I0vw0QG
=OUxo
-----END PGP SIGNATURE-----
--
Petr Novotny, ANTEK CS
[EMAIL PROTECTED]
http://www.antek.cz
PGP key ID: 0x3BA9BC3F
-- Don't you know there ain't no devil there's just God when he's drunk.
[Tom Waits]
>Adam D . McKenna writes:
> > | echo "Dear $SENDER,
> > Thank You for joining our e-newsletter. This e-mail is a confirmation
that
> > you have been added to our e-newsletter mailing list.
> > "
>
> > Is there a way to span lines like this?
>
>No. Either put it into a script, or put \n's into your echo.
>Personally, I think that a leading space following a program delivery
>ought to be appended to the command before it's handed to /bin/sh, but
>that's just me.
>
If you have to use \n just remember to use echo -e and not just echo. echo
without the -e will ignore the return sequence and just print \n
Cheers
Leon Vismer
> From: "Leon Vismer" <[EMAIL PROTECTED]>
> Date: Thu, 5 Aug 1999 17:18:59 +0200
>
> If you have to use \n just remember to use echo -e and not just echo. echo
> without the -e will ignore the return sequence and just print \n
This varies depending on your version of unix.
Chris
--
Chris Garrigues virCIO
http://www.DeepEddy.Com/~cwg/ http://www.virCIO.Com
+1 512 432 4046 +1 512 374 0500
4314 Avenue C
O- Austin, TX 78751-3709
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
Ah, the truth comes out...
Now does anyone know how to insert carriage returns under FreeBSD 3.0?
Dave
On Thursday, August 05, 1999 11:37 AM, Chris Garrigues
[SMTP:[EMAIL PROTECTED]] wrote:
> > From: "Leon Vismer" <[EMAIL PROTECTED]>
> > Date: Thu, 5 Aug 1999 17:18:59 +0200
> >
> > If you have to use \n just remember to use echo -e and not just echo. echo
> > without the -e will ignore the return sequence and just print \n
>
> This varies depending on your version of unix.
>
> Chris
>
> --
> Chris Garrigues virCIO
> http://www.DeepEddy.Com/~cwg/ http://www.virCIO.Com
> +1 512 432 4046 +1 512 374 0500
> 4314 Avenue C
> O- Austin, TX 78751-3709
>
>
> 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.
>
>
> << File: ATT00000.att >>
On Thu, Aug 05, 1999 at 12:21:58PM -0400, Dave Kitabjian wrote:
>
> Ah, the truth comes out...
>
> Now does anyone know how to insert carriage returns under FreeBSD 3.0?
What shell?
--
Brad Shelton On Line Exchange http://ole.net
Sorry, I forgot that it was builtin.
bourne shell is preferred, /bin/sh. But if I could even do it in csh, that would be
better than nothing!
Dave
On Thursday, August 05, 1999 12:26 PM, Brad Shelton [SMTP:[EMAIL PROTECTED]]
wrote:
> On Thu, Aug 05, 1999 at 12:21:58PM -0400, Dave Kitabjian wrote:
> >
> > Ah, the truth comes out...
> >
> > Now does anyone know how to insert carriage returns under FreeBSD 3.0?
>
> What shell?
>
> --
> Brad Shelton On Line Exchange http://ole.net
Dave Kitabjian ([EMAIL PROTECTED]) wrote:
: Ah, the truth comes out...
: Now does anyone know how to insert carriage returns under FreeBSD 3.0?
Two simple ways that work anywhere:
|echo "line one"; echo "line two"
|/usr/bin/printf 'line one\nline two\n'
-harold
On Wed, Aug 04, 1999 at 11:04:08PM -0400, Chris Johnson wrote:
> On Wed, Aug 04, 1999 at 08:15:06PM -0400, Brian Reichert wrote:
> > I must be utterly misunderstanding the documentation.
> >
> > I"m trying to create a system-wide alias to catch pager requests.
> >
> > So, I created:
> >
> > # cat ~alias/.qmail-page
> > | /usr/local/bin/sendpage -B -S -c Subject: -f "$EXT"
> >
> > # ls -l ~alias/.qmail-page
> > -rw-r--r-- 1 root qmail 84 Aug 4 19:55 /var/qmail/alias/.qmail-page
> Your ~alias/.qmail-page file should be ~alias/.qmail-page-default.
Yes, that did do it. I acutally tried that, and ran into an
unrelated probelm that led me to believe that was not working.
But I'm alive now. Once I get a properly fleshed-out invokation
of sendpage, I'll mention it on the list, as it looks handy for
people...
And thanks for the advice. I was in a blind rush, and was not
loking in the right place.
In part, I was assuming that qmail-getpw would return an error if
there was no valid mailbox for the localpart in question... Is
there a qmail utility that can emperically check whether a valid
mailbox exists (without, of course, sending mail)?
> Chris
--
Brian 'you Bastard' Reichert [EMAIL PROTECTED]
37 Crystal Ave. #303 Daytime number: (781) 899-7484 x704
Derry NH 03038-1713 USA Intel architecture: the left-hand path
Hi LDAP fans
The 19990805 release of the qmail-ldap integration patch is available
at http://www.nrg4u.com.
The target audience is ISP's, virtual domain users and LDAP freaks.
Changes:
- catchall account for domains now available
- new account status field (gives various grades to disable nonpaying
user's mail accounts)
- major rewrite of all qmail-local LDAP functions
- code review
- all reported bugs since the last release fixed (hopefully)
Have fun
--
Andre
FYI: MSX uses X.400 as the preferred connection between MSX hosts; but
this can be complicated to set up. SMTP works fine, and we also use a Unix
Mail Hub and MSX client-hosts.
I wonder if you might want to think about initiating the dial-up
connection from the main server to the satellite hosts, though? That way,
you could use 'serialmail' and virtual-hosting to queue each server's mail
in their own Maildirs, and a cron job to bring up the ISDN connections
one-by-one and blat the mail through. It would mean you wouldn't have to
work out how to get MSX to use ETRN (though I hear it's quite easy) as well.
By the way, qmail proper *doesn't* support ETRN, an add-on package called
'serialmail' can, instead. Install that and read through the man pages;
it's not hard to set it up.
-----Original Message-----
From: Olivier M. [mailto:[EMAIL PROTECTED]]
Sent: Sunday, August 01, 1999 11:28 PM
To: [EMAIL PROTECTED]
Subject: Re: M$ Exchange -> qmail
Thanks for your answer Thomas,
On Sun, Aug 01, 1999 at 11:47:27PM +0200, Thomas Neumann wrote:
> I'm not an expert on M$ Exchange by any means, but I know
> for sure that Exchange can run an SMTP server for itself,
> and this is where you can hook up to. Just have all customer
> mail delivered into Maildirs on your new qmail machine
> (one Maildir per customer), then you can use a small program
> to fetch the mail using POP3 from the qmail machine to the
> Windows machines running Exchange, where you re-inject them
> into the local SMTP server offered by Exchange. I do the
One maildir per user or per domain ? If it is per domain,
then it should be possible to use serialmail to send the mails
from the linux server to the NT client: it's good documented.
> I think Exchange can also use ETRN to tell another SMTP
> server that it wants it to send queued mail, but ETRN
> is even worse, being incredibly insecure and qmail doesn't
> support it w/o a patch anyway, so stay away from it.
okay... didn't found anything about ETRN on the qmail
homepage. then I'll make some tests using only smtp.
Thanks for your hints!
Olivier
When sending to a specific host I sometimes, but not always get this message
as a bounce
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]]
Sent: Sunday, August 01, 1999 12:35 AM
Subject: Status of: SPR entered into tracking system Ixxut not
[EMAIL PROTECTED]:]UCTION SOURCE!/2/99 9:50 AM
SMTP/X400 conversion failure
I'm 90% sure this is nothing on my end but something on theirs.
Can anyone give me insight into this?
Tim Hunter
[EMAIL PROTECTED]
CIMx Company
1001 Ford Circle
Cincinnati, OH 45150
p 513 248-7700
f 513-248-7711
http://www.cimx.com
On Thu, 5 Aug 1999, Tim Hunter wrote:
> When sending to a specific host I sometimes, but not always get this message
> as a bounce
>
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]
> Sent: Sunday, August 01, 1999 12:35 AM
> Subject: Status of: SPR entered into tracking system Ixxut not
> [EMAIL PROTECTED]:]UCTION SOURCE!/2/99 9:50 AM
>
>
> SMTP/X400 conversion failure
>
> I'm 90% sure this is nothing on my end but something on theirs.
> Can anyone give me insight into this?
You're going to have to give some more information on the message being
sent... It's about 5 years since i struggled with X.400<>RFC822 based
mail.
I suspect that the message being received (I don't even know if it's
being received as SMTP or X.400 from your message) is either a MIME
message or a X.400 message with a complex body-part.
The gateway is then converting the complex types into the corresponding
complex type, in otherwised it might be taking a P22 body-part and
splitting it into a couple of IA5 and an MS-WORD part, and then
recombining these to make a MIME message.
The difficulty happens when the gateway doing the conversion doesn't have
an appropraite filter installed, and the software authors haven't
considered providing a default encoder. this is partiticulary difficult
with x400->SMTP make since the body-part-types are tagged with OIDS which
might be specific to the site, the application, or anything one likes...
basically if you want to pass complex documents between X.400 and MIME
then you need someone very clueful running the realy system who is
prepared to add extra encoders as required.
Richard
Hello folks,
maybe this question was answered before, but I didn�t find it in the
archive, so
there we go:
I want to setup qmail in a chroot environment (qmail won�t run on the
main system
but in this virtual machine) ...
The Problem is, that qmail-smtpd receives the mails (they�re stored in
the mail queue) but they
aren�t delivered to the users maildir ... also there is no entry in the
logfile; sometimes the mail
will be delivered after 10 minutes or so.
Some people told me thats because of an DNS problem, but this can�t be
the reason because it is
already in the mail queue directory.
Any guesses ?
Regards,
Joerg Staedele
On Thu, Aug 05, 1999 at 07:42:32PM +0200, J�rg St�dele wrote:
> Hello folks,
>
> maybe this question was answered before, but I didn�t find it in the
> archive, so
> there we go:
>
> I want to setup qmail in a chroot environment (qmail won�t run on the
> main system
> but in this virtual machine) ...
>
> The Problem is, that qmail-smtpd receives the mails (they�re stored in
> the mail queue) but they
> aren�t delivered to the users maildir ... also there is no entry in the
> logfile; sometimes the mail
> will be delivered after 10 minutes or so.
>
> Some people told me thats because of an DNS problem, but this can�t be
> the reason because it is
> already in the mail queue directory.
[root@koek] ~# ls -al /var/qmail/queue/lock/
total 3
drwxr-x--- 2 qmailq qmail 1024 Jul 5 18:27 ./
drwxr-x--- 11 qmailq qmail 1024 Jul 5 18:27 ../
-rw------- 1 qmails qmail 0 Jul 5 18:27 sendmutex
-rw-r--r-- 1 qmailr qmail 1024 Aug 4 13:38 tcpto
prw--w--w- 1 qmails qmail 0 Aug 5 22:14 trigger|
A delay in local delivery usually occurs when qmail-queue can't find
/var/qmail/queue/lock/trigger. Have a look at that :)
Greetz, Peter
--
| 'He broke my heart, | Peter van Dijk |
I broke his neck' | [EMAIL PROTECTED] |
nognikz - As the sun | Hardbeat@ircnet - #cistron/#linux.nl |
http://www.nognikz.mdk.nu/ | Hardbeat@undernet - #groningen/#kinkfm/#vdh |
My problem is specific to qmail-1.03 patched with jbuce. It seems to be
rejecting mail where the sender domain has a valid MX record but no IP
address. (One example is hotoffice.com.) Can anyone say what may be
happening? Are there better patches I could be using to deter spam,
increase logging, &c.?
Thanks for any help/pointers!
/pg
--
Peter Green
Gospel Communications Network, SysAdmin
[EMAIL PROTECTED]
Quoting Peter Green ([EMAIL PROTECTED]):
> My problem is specific to qmail-1.03 patched with jbuce. It seems to be
> rejecting mail where the sender domain has a valid MX record but no IP
> address. (One example is hotoffice.com.) Can anyone say what may be
> happening? Are there better patches I could be using to deter spam,
> increase logging, &c.?
We're using 1.03 and the jbuce patch, and hotoffice.com is accepted
in the return path on our machine.
Aaron
220 cois.coinet.com ESMTP
helo localhost
250 cois.coinet.com
mail from:<[EMAIL PROTECTED]>
250 ok
rcpt to:<[EMAIL PROTECTED]>
250 ok
On Thu, 5 Aug 1999, Aaron L. Meehan wrote:
> We're using 1.03 and the jbuce patch, and hotoffice.com is accepted
> in the return path on our machine.
That being the case, is there a reasonable explanation for the following
entry in our maillog? (Usernames and local domain omitted.)
Aug 5 10:29:08 joppa qmail-smtpd: MAIL FROM MX check failed
(****@hotoffice.com) -> (****@****.***) [207.69.200.32] (HELO
smtp2.mindspring.com)
Other mail is getting through just fine for this local user...any
thoughts?
/pg
--
Peter Green
Gospel Communications Network, SysAdmin
[EMAIL PROTECTED]
On Thu, Aug 05, 1999 at 02:55:17PM -0400, Peter Green wrote:
> On Thu, 5 Aug 1999, Aaron L. Meehan wrote:
> > We're using 1.03 and the jbuce patch, and hotoffice.com is accepted
> > in the return path on our machine.
>
> That being the case, is there a reasonable explanation for the following
> entry in our maillog? (Usernames and local domain omitted.)
>
> Aug 5 10:29:08 joppa qmail-smtpd: MAIL FROM MX check failed
> (****@hotoffice.com) -> (****@****.***) [207.69.200.32] (HELO
> smtp2.mindspring.com)
>
> Other mail is getting through just fine for this local user...any
> thoughts?
I know what the problem is.. It's because of the **** ***** ** ***** * ***
** *****.
--Adam
> /pg
> --
> Peter Green
> Gospel Communications Network, SysAdmin
> [EMAIL PROTECTED]
>
Quoting Peter Green ([EMAIL PROTECTED]):
> On Thu, 5 Aug 1999, Aaron L. Meehan wrote:
> > We're using 1.03 and the jbuce patch, and hotoffice.com is accepted
> > in the return path on our machine.
>
> That being the case, is there a reasonable explanation for the following
> entry in our maillog? (Usernames and local domain omitted.)
>
> Aug 5 10:29:08 joppa qmail-smtpd: MAIL FROM MX check failed
> (****@hotoffice.com) -> (****@****.***) [207.69.200.32] (HELO
> smtp2.mindspring.com)
Perhaps you are running a different revision of the patch? The patch
I have lists it as rev C, and messages to syslog on my machine don't
say "MAIL FROM MX check failed", rather "MAIL FROM DNS check failed",
so I'm tempted to believe that we have different patches.
Aaron
On Thu, 5 Aug 1999, Adam D . McKenna wrote:
> I know what the problem is.. It's because of the **** ***** ** ***** * ***
> ** *****.
Of course! What I *meant* to say was that the log entry looks like:
Aug 5 10:29:08 joppa qmail-smtpd: MAIL FROM MX check failed
([EMAIL PROTECTED]) -> ([EMAIL PROTECTED]) [207.69.200.32] (HELO
smtp2.mindspring.com)
My apologies...(thanks Adam)
/pg
--
Peter Green
Gospel Communications Network, SysAdmin
[EMAIL PROTECTED]
> Linux has a similar thing in the /proc filesystem, as someone else has
> already pointed out.
If you're interested in doing this on Linux, go read
/usr/src/linux/Documentation/proc.txt. You can set the maximum number of
file descriptors for the system (the default is 4096), but to increase the
max per process (I think that's right) you have to modify the kernel (the
default is 1024). Instructions are in that file (look for file-nr and
file-max).
Troy
I'm having trouble when using qmail to relay mail via smtp. When I send email to an
address like
[EMAIL PROTECTED], I get a failire notice which says:
<[EMAIL PROTECTED]"">:
Sorry, I couldn't find any host named harmony-ds.com"". (#5.1.2)
Note the two quotes on the end of the host name.
This happens when sending mail to any address from a host other than the host running
qmail.
Although, strangely enough, it does not happen when sending mail via Netscape
Messenger
running on MacOS. It does happen with Netscape on linux and Windows.
any ides? I'm not on this list, so please reply to [EMAIL PROTECTED]
thanks,
- Scott
Hi,
I'm experimenting with qmail's throughput. It approximates what
we want to do with it. (send mail from a mail database through qmail-inject)
I want to see how fast qmail will queue messages.
I have a ~433 Mhz Alpha.
Experiment:
script that runs qmail-inject 1000 times
with a 1000 byte body
Qmail-send is not running
Result:
takes 73 seconds (13.7/sec)
This seems a little slow to me. The system cpu is 70% idle.
Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
Is this normal ? Any ideas on how to speed it up.
thanks,
-jim
There is something very wrong here. Cat'ing two blocks of data should
take milliseconds. Are you out of memory (does vmstat show paging?)?
When your cat is << .1 seconds or so then rerun your tests. Wait a
minute, did you run the cat WHILE you were trying to do the deliveries?
If so, then that is about right as qmail is beating on your hard disk.
Your qmail-inject seem like they are off by about a factor of 10 or less
(since qmail-send isn't running), but that is just a swag (assuming
about a hundred other things like a properly configured file system, no
swapping/page stealing (sufficent memory), and a current-technology SCSI
non-IDE drive).
Jim Arnott wrote:
>
> Hi,
>
> I'm experimenting with qmail's throughput. It approximates what
> we want to do with it. (send mail from a mail database through qmail-inject)
> I want to see how fast qmail will queue messages.
>
> I have a ~433 Mhz Alpha.
>
> Experiment:
>
> script that runs qmail-inject 1000 times
> with a 1000 byte body
> Qmail-send is not running
>
> Result:
> takes 73 seconds (13.7/sec)
>
> This seems a little slow to me. The system cpu is 70% idle.
>
> Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
>
> Is this normal ? Any ideas on how to speed it up.
>
> thanks,
> -jim
--
Daemeon Reiydelle
Systems Engineer, Anthropomorphics Inc.
[EMAIL PROTECTED]
At 01:56 PM Thursday 8/5/99, Jim Arnott wrote:
>Hi,
>
>I'm experimenting with qmail's throughput. It approximates what
>we want to do with it. (send mail from a mail database through qmail-inject)
>I want to see how fast qmail will queue messages.
>
>I have a ~433 Mhz Alpha.
>
>Experiment:
>
> script that runs qmail-inject 1000 times
> with a 1000 byte body
> Qmail-send is not running
>
>Result:
> takes 73 seconds (13.7/sec)
>
>This seems a little slow to me. The system cpu is 70% idle.
>
>Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
>
>
>Is this normal ? Any ideas on how to speed it up.
First question. Did you try the experiment:
cat >/dev/null <100byte.in ?
Second question. Why do you think that the cat command is comparable to
inserting a structured object into a sophisticated queue system that ensures
reliability performance at high numbers?
If this is a rhetorical question, then what is your expectation of the cost
of queuing a mail message and on what basis did you make that expectation?
Mark.
At 12:28 PM Thursday 8/5/99, Daemeon Reiydelle wrote:
>There is something very wrong here. Cat'ing two blocks of data should
>take milliseconds. Are you out of memory (does vmstat show paging?)?
I though he mentioned that he ran it 1000 times and thus the numbers reflect
a 1000 invocations...
>When your cat is << .1 seconds or so then rerun your tests. Wait a
>minute, did you run the cat WHILE you were trying to do the deliveries?
>If so, then that is about right as qmail is beating on your hard disk.
>
>Your qmail-inject seem like they are off by about a factor of 10 or less
>(since qmail-send isn't running), but that is just a swag (assuming
>about a hundred other things like a properly configured file system, no
>swapping/page stealing (sufficent memory), and a current-technology SCSI
>non-IDE drive).
>
>Jim Arnott wrote:
> >
> > Hi,
> >
> > I'm experimenting with qmail's throughput. It approximates what
> > we want to do with it. (send mail from a mail database through
> qmail-inject)
> > I want to see how fast qmail will queue messages.
> >
> > I have a ~433 Mhz Alpha.
> >
> > Experiment:
> >
> > script that runs qmail-inject 1000 times
> > with a 1000 byte body
> > Qmail-send is not running
> >
> > Result:
> > takes 73 seconds (13.7/sec)
> >
> > This seems a little slow to me. The system cpu is 70% idle.
> >
> > Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
> >
> > Is this normal ? Any ideas on how to speed it up.
> >
> > thanks,
> > -jim
>
>--
>Daemeon Reiydelle
>Systems Engineer, Anthropomorphics Inc.
>[EMAIL PROTECTED]
Hi,
I meant the "cat >> out.file < 1000byte.in" was done in a 1000 time loop.
#vmstat 1 (during inject test):
Virtual Memory Statistics: (pagesize = 8192)
procs memory pages intr cpu
r w u act free wire fault cow zero react pin pout in sy cs us sy id
2120 29 8705 18K 4026 1141 453 343 0 382 0 118 3K 1K 2 24 74
2120 31 8790 18K 3977 670 260 216 0 224 0 125 2K 1K 2 15 83
4119 29 8822 18K 3935 989 395 291 0 321 0 96 3K 1K 4 27 69
2120 30 8825 18K 3936 1030 418 309 0 375 0 118 3K 1K 2 27 71
2120 31 8856 18K 3953 1049 426 309 0 360 0 109 4K 1K 4 29 67
2120 30 8833 18K 3928 1231 501 354 0 436 0 121 2K 1K 3 22 75
4119 29 8836 18K 3932 667 258 216 0 221 0 120 2K 1K 2 18 80
256 Meg mem, Latest 10 Gig SCSI
-jim
>
> There is something very wrong here. Cat'ing two blocks of data should
> take milliseconds. Are you out of memory (does vmstat show paging?)?
> When your cat is << .1 seconds or so then rerun your tests. Wait a
> minute, did you run the cat WHILE you were trying to do the deliveries?
> If so, then that is about right as qmail is beating on your hard disk.
>
> Your qmail-inject seem like they are off by about a factor of 10 or less
> (since qmail-send isn't running), but that is just a swag (assuming
> about a hundred other things like a properly configured file system, no
> swapping/page stealing (sufficent memory), and a current-technology SCSI
> non-IDE drive).
>
> Jim Arnott wrote:
> >
> > Hi,
> >
> > I'm experimenting with qmail's throughput. It approximates what
> > we want to do with it. (send mail from a mail database through qmail-inject)
> > I want to see how fast qmail will queue messages.
> >
> > I have a ~433 Mhz Alpha.
> >
> > Experiment:
> >
> > script that runs qmail-inject 1000 times
> > with a 1000 byte body
> > Qmail-send is not running
> >
> > Result:
> > takes 73 seconds (13.7/sec)
> >
> > This seems a little slow to me. The system cpu is 70% idle.
> >
> > Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
> >
> > Is this normal ? Any ideas on how to speed it up.
> >
> > thanks,
> > -jim
>
> --
> Daemeon Reiydelle
> Systems Engineer, Anthropomorphics Inc.
> [EMAIL PROTECTED]
>
> At 01:56 PM Thursday 8/5/99, Jim Arnott wrote:
>
>
> >Hi,
> >
> >I'm experimenting with qmail's throughput. It approximates what
> >we want to do with it. (send mail from a mail database through qmail-inject)
> >I want to see how fast qmail will queue messages.
> >
> >I have a ~433 Mhz Alpha.
> >
> >Experiment:
> >
> > script that runs qmail-inject 1000 times
> > with a 1000 byte body
> > Qmail-send is not running
> >
> >Result:
> > takes 73 seconds (13.7/sec)
> >
> >This seems a little slow to me. The system cpu is 70% idle.
> >
> >Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
> >
> >
> >Is this normal ? Any ideas on how to speed it up.
>
>
> First question. Did you try the experiment:
>
> cat >/dev/null <100byte.in ?
that takes 3 seconds.
BTW thats 1000 bytes.
>
>
> Second question. Why do you think that the cat command is comparable to
> inserting a structured object into a sophisticated queue system that ensures
> reliability performance at high numbers?
Show the absolute best test scenario and to demonstrate that my system is at
least working.
> If this is a rhetorical question, then what is your expectation of the cost
> of queuing a mail message and on what basis did you make that expectation?
I'm just trying to figure out what kind of throughput to expect and justify
that against what managment expects.
IMHO it seems that there is somthing strange going on when while doing the
inject test the CPU is 70% idle. (cat test is 0% idle). There seems
to be a bottleneck somewhere.
-jim
At 03:53 PM Thursday 8/5/99, Jim Arnott wrote:
> > >I have a ~433 Mhz Alpha.
> > >
> > >Experiment:
> > >
> > > script that runs qmail-inject 1000 times
> > > with a 1000 byte body
> > > Qmail-send is not running
> > >
> > >Result:
> > > takes 73 seconds (13.7/sec)
> > >
> > >This seems a little slow to me. The system cpu is 70% idle.
> > >
> > >Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
IMHO it seems that there is somthing strange going on when while doing the
>inject test the CPU is 70% idle. (cat test is 0% idle). There seems
>to be a bottleneck somewhere.
Their is *always* a "bottleneck" in every test. A bottleneck is absolutely
normal.
In our line of work, it is typically a cpu or or disk bottleneck, depending
on the nature of the program involved.
I think you've discovered that your "cat" test has a cpu bottleneck and the
qmail-inject test has a disk bottleneck. I don't think you can conclude
anything more than that from your test.
Mark.
Jim Arnott <[EMAIL PROTECTED]> writes on 5 August 1999 at 15:53:53 -0500
> IMHO it seems that there is somthing strange going on when while doing the
> inject test the CPU is 70% idle. (cat test is 0% idle). There seems
> to be a bottleneck somewhere.
That's fairly normal; queue disk bandwidth is generally the bottleneck
on massive inject tests.
A lot of strategies for increasing bandwidth for massing mailings
amount to various ways to get around the queue.
One way is to decide that you don't care if a few messages are lost or
duplicated if you have a crash during the injection phase. In that
case you can hack qmail to not sync after queueing a message -- makes
a *huge* difference, but of course your queue isn't safe from a
crash.
Holding the queue in a battery-backed ramdisk of suitable size is a
neat compromise -- very fast, and fairly stable.
Another way is to make an initial attempt to deliver the message
directly, before queueing it. Since some fairly large proportion of
messages go through on the first try (depends a lot on your address
mix, but 50% to 95%), queueing only the ones that fail of initial
delivery can also be a big win.
--
David Dyer-Bennet ***NOTE ADDRESS CHANGES*** [EMAIL PROTECTED]
http://dd-b.lighthunters.net/ (photos) Minicon: http://www.mnstf.org/minicon
http://www.dd-b.net/dd-b (sf) http://ouroboros.demesne.com/ Ouroboros Bookworms
Join the 20th century before it's too late!
[snip]
> Their is *always* a "bottleneck" in every test. A bottleneck is absolutely
> normal.
>
> In our line of work, it is typically a cpu or or disk bottleneck, depending
> on the nature of the program involved.
[snip]
You have to learn to use your systems performance monitering tools AND
how to interpret their reports. different system shave different tools.
SYSV UNIX has the sar command, Linux has top, vmstat, and others.
--
Richard Shetron [EMAIL PROTECTED] [EMAIL PROTECTED]
What is the Meaning of Life?
There is no meaning,
It's just a consequence of complex carbon based chemistry; don't worry about it
The Super 76, "Free Aspirin and Tender Sympathy", Las Vegas Strip.
Please review the man page for vmstats. You may find it helpful to buy
and read any of the books on UNIX performance tuning (the ones on
Solaris tuning would be most usefull). Your swap file is in serious
contention with your processes since you are badly over committed. You
might also want to reduce the number of processes doing qmail injection.
You will find your performance improves dramatically.
Jim Arnott wrote:
>
> Hi,
>
> I meant the "cat >> out.file < 1000byte.in" was done in a 1000 time loop.
>
> #vmstat 1 (during inject test):
>
> Virtual Memory Statistics: (pagesize = 8192)
> procs memory pages intr cpu
> r w u act free wire fault cow zero react pin pout in sy cs us sy id
> 2120 29 8705 18K 4026 1141 453 343 0 382 0 118 3K 1K 2 24 74
> 2120 31 8790 18K 3977 670 260 216 0 224 0 125 2K 1K 2 15 83
> 4119 29 8822 18K 3935 989 395 291 0 321 0 96 3K 1K 4 27 69
> 2120 30 8825 18K 3936 1030 418 309 0 375 0 118 3K 1K 2 27 71
> 2120 31 8856 18K 3953 1049 426 309 0 360 0 109 4K 1K 4 29 67
> 2120 30 8833 18K 3928 1231 501 354 0 436 0 121 2K 1K 3 22 75
> 4119 29 8836 18K 3932 667 258 216 0 221 0 120 2K 1K 2 18 80
>
> 256 Meg mem, Latest 10 Gig SCSI
>
> -jim
>
> >
> > There is something very wrong here. Cat'ing two blocks of data should
> > take milliseconds. Are you out of memory (does vmstat show paging?)?
> > When your cat is << .1 seconds or so then rerun your tests. Wait a
> > minute, did you run the cat WHILE you were trying to do the deliveries?
> > If so, then that is about right as qmail is beating on your hard disk.
> >
> > Your qmail-inject seem like they are off by about a factor of 10 or less
> > (since qmail-send isn't running), but that is just a swag (assuming
> > about a hundred other things like a properly configured file system, no
> > swapping/page stealing (sufficent memory), and a current-technology SCSI
> > non-IDE drive).
> >
> > Jim Arnott wrote:
> > >
> > > Hi,
> > >
> > > I'm experimenting with qmail's throughput. It approximates what
> > > we want to do with it. (send mail from a mail database through qmail-inject)
> > > I want to see how fast qmail will queue messages.
> > >
> > > I have a ~433 Mhz Alpha.
> > >
> > > Experiment:
> > >
> > > script that runs qmail-inject 1000 times
> > > with a 1000 byte body
> > > Qmail-send is not running
> > >
> > > Result:
> > > takes 73 seconds (13.7/sec)
> > >
> > > This seems a little slow to me. The system cpu is 70% idle.
> > >
> > > Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
> > >
> > > Is this normal ? Any ideas on how to speed it up.
> > >
> > > thanks,
> > > -jim
> >
> > --
> > Daemeon Reiydelle
> > Systems Engineer, Anthropomorphics Inc.
> > [EMAIL PROTECTED]
--
Daemeon Reiydelle
Systems Engineer, Anthropomorphics Inc.
[EMAIL PROTECTED]
On Thu, Aug 05, 1999 at 01:56:02PM -0500, Jim Arnott wrote:
> Experiment:
>
> script that runs qmail-inject 1000 times
> with a 1000 byte body
> Qmail-send is not running
>
> Result:
> takes 73 seconds (13.7/sec)
>
> This seems a little slow to me. The system cpu is 70% idle.
>
> Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
The reason this is slow is that large queue injections are very disk
intensive. To understand why, read the INTERNALS document. This will
help you understand the difference between what qmail-inject does and
what cat does.
To solve this problem, use a SCSI controller which has a write-back
cache. Generally, you have to look at low-end RAID controllers to
get this.
If you're sending idendical messages, it's better to use one qmail-inject
with many recipients.
--
John White johnjohn
at
triceratops.com
PGP Public Key: http://www.triceratops.com/john/public-key.pgp
On Thu, Aug 05, 1999 at 01:56:02PM -0500, Jim Arnott wrote:
> Experiment:
>
> script that runs qmail-inject 1000 times
> with a 1000 byte body
> Qmail-send is not running
>
> Result:
> takes 73 seconds (13.7/sec)
>
> This seems a little slow to me. The system cpu is 70% idle.
>
> Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
Since you have a custom application, consider running
qmail-queue directly rather than going through qmail-inject.
-harold
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
> Experiment:
>
> script that runs qmail-inject 1000 times
> with a 1000 byte body
> Qmail-send is not running
>
> Result:
> takes 73 seconds (13.7/sec)
>
> This seems a little slow to me. The system cpu is 70% idle.
What processes are taking those 30%? Isn't it syslogd?
> Same experiment with "cat >> out.file < 1000byte.in" takes 4 seconds.
Does this cat sync after each file write, append a few lines to a log
and sync again?
-----BEGIN PGP SIGNATURE-----
Version: PGP 6.0.2 -- QDPGP 2.60
Comment: http://community.wow.net/grt/qdpgp.html
iQA/AwUBN6qVM1MwP8g7qbw/EQLGrgCg0ID695PcKg7PbBbEvPLmezaL4DQAoOvw
R6UXMgRpoCfjgq6ZHNWP0BqO
=Ou9M
-----END PGP SIGNATURE-----
--
Petr Novotny, ANTEK CS
[EMAIL PROTECTED]
http://www.antek.cz
PGP key ID: 0x3BA9BC3F
-- Don't you know there ain't no devil there's just God when he's drunk.
[Tom Waits]
Just to let everyone know how easy it really was (especially once
I got pointers about correctly naming the .qmail file):
To create system-wide mailbox to catch pager requests:
# cat ~alias/.qmail-page-default
# Attempt to send a page to sendpage. Assumes sendpage recipient
# contains no hypens.
# Bounce if this is not a valid pager recipient.
| echo $EXT2 | /usr/local/bin/sendpage -bv | grep 'no expansion' && echo
"invalid pager recipient: ($EXT2)" && exit 100 ; exit 0
# Otherwise, send off to sendpage.
| /usr/local/bin/sendpage -B -S -c Subject: $EXT2
(Long lines do not wrap.)
Now <[EMAIL PROTECTED]> works in a useful way...
Three cheers for qmail for making it simple! And I hate Bourne
Shell! Thank you very much...
--
Brian 'you Bastard' Reichert [EMAIL PROTECTED]
37 Crystal Ave. #303 Daytime number: (781) 899-7484 x704
Derry NH 03038-1713 USA Intel architecture: the left-hand path
Hi there :)
I've just been reading over the docs for qmail. It says to HUP the server if
you change any files in the /var/qmail/control directory..
What is a safe way to do this?
I've got a /etc/init.d/qmail script which doesn't actually work. I know that
I have to HUP the tcpserver processes and the qmail-smtpd processes, but
does anyone know exactly what to do?
Any help is most appreciated :)
Regards,
Marc-Adrian Napoli
Connect Infobahn Australia
+61 2 92811750
On Fri, 6 Aug 1999, Marc-Adrian Napoli wrote:
> I've just been reading over the docs for qmail. It says to HUP the server if
> you change any files in the /var/qmail/control directory..
Actually, only locals and virtualdomains; see qmail-send(8). For the
other files read on startup by the long-running daemons you need to shut
down and restart qmail-send:
CONTROL FILES
WARNING: qmail-send reads its control files only when it
starts. If you change the control files, you must stop
and restart qmail-send. Exception: If qmail-send receives
a HUP signal, it will reread locals and virtualdomains.
> What is a safe way to do this?
See the admin FAQ: ftp://koobera.math.uic.edu/www/qmail/faq/admin.html
> I've got a /etc/init.d/qmail script which doesn't actually work. I know that
> I have to HUP the tcpserver processes and the qmail-smtpd processes, but
> does anyone know exactly what to do?
qmail-smtpd runs fresh for each new connection, and tcpserver doesn't read
any control files. The only long-running part of qmail you need to worry
about notifying or restarting is qmail-send (it takes care of it's support
processes (qmail-clean and qmail-[lr]spawn) for you).
-m
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
> I've just been reading over the docs for qmail. It says to HUP the server
> if you change any files in the /var/qmail/control directory..
>
> What is a safe way to do this?
If you're running BSD-based system (linux, *BSD)
killall -HUP qmail-send
is what you need. If you're running SysV-based system, don't ever
try that and do
kill -SIGHUP `pidof qmail-send`
instead.
> I've got a /etc/init.d/qmail script which doesn't actually work. I know
> that I have to HUP the tcpserver processes and the qmail-smtpd processes,
> but does anyone know exactly what to do?
You don't HUP qmail-smtpd - it gets respawned for every
connection. And you don't HUP tcpeserver - the only config it uses
is that -x something.cdb file and if gets re-read everytime a
connection arrives.
You only HUP qmail-send.
-----BEGIN PGP SIGNATURE-----
Version: PGP 6.0.2 -- QDPGP 2.60
Comment: http://community.wow.net/grt/qdpgp.html
iQA/AwUBN6qYulMwP8g7qbw/EQLyYQCdFsWyVaV1fVJNw7gQfmmQy2IsPPgAoI+Q
Ry936cp/5GX0JEi3X3RTNKa7
=QiXN
-----END PGP SIGNATURE-----
--
Petr Novotny, ANTEK CS
[EMAIL PROTECTED]
http://www.antek.cz
PGP key ID: 0x3BA9BC3F
-- Don't you know there ain't no devil there's just God when he's drunk.
[Tom Waits]
Hello,
I have a very interesting situation here. I have two servers,
host1.mydomain.com and host2.mydomain.com. I have host1 as the primary
mail server in the DNS-MX record. Both the hosts have qmail-1.03
installed in it.
I have user mailboxes in both the server:
user1, user2, user3 in host1
user4, user5, user6 in host2.
Now i would like to make qmail in my whole system work in such a way
that any mail to [EMAIL PROTECTED] or [EMAIL PROTECTED] or
[EMAIL PROTECTED] be delivered to user1, user2, user3 in HOST1
whereas mail to [EMAIL PROTECTED] [EMAIL PROTECTED]
[EMAIL PROTECTED] be delivered to mailboxes user4, user5, user6 in
HOST2.
Later i would like to implement these concepts to my hosts in
different cities (POP) so that every user of our network will have
same template of e-mail address ([EMAIL PROTECTED]) without any
mention of the hostname of the server in his/her city.
If anybody has any idea making this distributed system, please help.
Thanks in advance.
Jason
