Analysis of reasons for bounce responses
![](https://secure.gravatar.com/avatar/632bc87a86b0db6153c56defac0ba682.jpg?s=120&d=mm&r=g)
This is not my problem but one soemone approached me about off list.
I'll summarise their situation and I would appreciate any thoughts,
comments, sanity checking on my analysis.
Situation:
List admin's domain is hosted by a server run by a hosting company.
The server is (probably) running Linux
Mailman 2.1.2 is provided on the server with, I believe, CPanel
The server's MTA appears to be Exim 4.20 handling both incoming and
outgoing messages to/from MailmanMailman list's basically work OK much of the time
Periodically messages out from Mailman are being bounced en masse
leading to the subscriber accounts being disabledThe bounce response attached to the bounce action notice produced by
Mailman when subscribers are disabled in these circumstances is from
the server local MTA. The MTA lists what appears to be all the
addresses to which the outgoing message was sent with a note against
each saying "unrouteable mail domain maildomain.tld". The bounce
response asserts that this is a permanent failure. The messages are
each being sent to a number of different mail domains not just a single
oneThe list admin with the problem has no access to either the Mailman
or MTA logsThe server administrator is supposedly claiming that these bounces
are a result of "a bug in Mailman for which no patch is available"
My analysis is:
The problem is not obviously due to any bug in Mailman
The problem is more likely due to an IP routing problem which
periodically afflicts the server and its ability to reach the MTAs for
the mail domains addressed by the outgoing messagesThe MTA is accepting the outgoing messages from Mailman and then
finding it has problems of its own in delivering mail which are
unrelated to Mailman as the origin of the messages
But I have to concede I know nothing about Exim 4.20. I know even less
about CPanel.
Any thoughts or anatomically feasible suggestions welcome.
Richard Barrett
http://www.openinfo.co.uk
![](https://secure.gravatar.com/avatar/d4258e6caa5b231603e992b430199997.jpg?s=120&d=mm&r=g)
On Thu, 2003-07-31 at 06:08, Richard Barrett wrote:
I have seen this behavior a few times and in each case it was resource problem on the server. They need to check on the memory of the box and how much is being used during the bounces. I'm afraid it is only something that the sysadmin can really do.
I have also seen this problem when the /var directory was made too small. Could this be a problem with the users services being chrooted and them running out of quota space while processing large messages?
Good Luck - Jon Carnes
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
This is the EXACT problem I am having with mailman, using cpanel and exim. I, however, am on a dedicated box and have root access. I have been working on this problem for over three weeks. I am tired and frustrated and wishing I knew a little more about mailman and its inner workings. I come from a majordomo background. When I got this dedicated server, it had mailman pre-installed so I thought kewl I will use it instead. It ran fabulously until cpanel did their last upgrade. I have been in hell since then.
I wish I could help you, but I have not been able to figure out even from the logs what the issues are. Granted, looking at the logs are like gobbledigook to me. So I have spent a lot of time searching the archives for this problem. Although, I haven't been able to find anyone with my specific issue. I can tell you that qrunner has been consuming my cpu at astonishing rates. But there is no rhyme or reason to that, that I can see. If I find anything out I will post it here.
Chris
-----Original Message----- From: mailman-users-bounces+chris=christinedelarosa.com@python.org [mailto:mailman-users-bounces+chris=christinedelarosa.com@python.org] On Behalf Of Richard Barrett Sent: Thursday, July 31, 2003 5:08 AM To: Mailman Subject: [Mailman-Users] Analysis of reasons for bounce responses
This is not my problem but one soemone approached me about off list.
I'll summarise their situation and I would appreciate any thoughts,
comments, sanity checking on my analysis.
Situation:
List admin's domain is hosted by a server run by a hosting company.
The server is (probably) running Linux
Mailman 2.1.2 is provided on the server with, I believe, CPanel
The server's MTA appears to be Exim 4.20 handling both incoming and
outgoing messages to/from MailmanMailman list's basically work OK much of the time
Periodically messages out from Mailman are being bounced en masse
leading to the subscriber accounts being disabledThe bounce response attached to the bounce action notice produced by
Mailman when subscribers are disabled in these circumstances is from
the server local MTA. The MTA lists what appears to be all the
addresses to which the outgoing message was sent with a note against
each saying "unrouteable mail domain maildomain.tld". The bounce
response asserts that this is a permanent failure. The messages are
each being sent to a number of different mail domains not just a single
one
The list admin with the problem has no access to either the Mailman
or MTA logsThe server administrator is supposedly claiming that these bounces
are a result of "a bug in Mailman for which no patch is available"
My analysis is:
The problem is not obviously due to any bug in Mailman
The problem is more likely due to an IP routing problem which
periodically afflicts the server and its ability to reach the MTAs for
the mail domains addressed by the outgoing messagesThe MTA is accepting the outgoing messages from Mailman and then
finding it has problems of its own in delivering mail which are
unrelated to Mailman as the origin of the messages
But I have to concede I know nothing about Exim 4.20. I know even less
about CPanel.
Any thoughts or anatomically feasible suggestions welcome.
Richard Barrett
http://www.openinfo.co.uk
Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
This message was sent to: chris@christinedelarosa.com Unsubscribe or change your options at http://mail.python.org/mailman/options/mailman-users/chris%40christinede larosa.com
![](https://secure.gravatar.com/avatar/632bc87a86b0db6153c56defac0ba682.jpg?s=120&d=mm&r=g)
Chris
On Friday, August 1, 2003, at 12:48 am, Christine De La Rosa wrote:
As I said in my post, this bounce issue is not, in my view, at heart a
bug in Mailman. The evidence I saw of bounce notices from local MTA
(Exim) say that the MTA is having problems getting outgoing mail off
the system and finally bouncing it back to Mailman when it fails to do
so.
If you are getting final bounce notices from Mailman when it disables a
subscriber because of these problems what is the reason reported in the
bounce message from the MTA, which should be attached to the bottom of
the final bounce notification from MM? In the example final bounce
notice passed to me for comment the MTA was reporting it was having an
IP routing problem; that is hardly a Mailman generated problem as far
as I can tell.
One of my problems with looking at this is I am trying to help out a
fairly new user with a problem I have not seen on my systems, while I
have no access to the Mailman or MTA logs on a system experiencing the
problem. I have no standing with the CPanel owners to raise matters
with them in case they are contributing to the problem with the way
they set things up. Deeply frustrating.
Let me know if you gain any insight or information related to this.
Mailman can consume large amounts of CPU when the MTA it is trying to
hand mail over to repeatedly returns temporary failures. MM can be too
aggressive in repeatedly trying to get the MTA to either accept or
permanently fail the handover. There are changes in the Mailman CVS to
moderate MM's efforts in this respect but so far as I know those
changes are not available as a patch for MM 2.1.2 stable. This issue
may be affecting you but I am no expert myself in setting up Exim so as
to avoid this undesirable interaction; I do not see it with my Sendmail
setup. There have been posts to mailman-users on the topic of high CPU
utilisation. This one fromJon Carnes relates to Postfix MTA but its
substance/recommendations may be applicable to Exim:
http://www.mail-archive.com/mailman-users@python.org/msg17786.html
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
Hi Richard,
Here is what I know so far, mail for most list members seem to be going through for the most part. About two weeks ago MM mass unscribed about 300 folx. Now we seem to have stablized the members but the owners of the lists are getting routinely bounced. Yahoo email address do not have a problem, aol is a major headache as is hotmail. What is happening on my side is that the sub notices are never reaching the owners, so they are bouncing like mad and eventually hit the excessive bounce phase after five days (since we have the settings set to set to nomail after 5 bounces and mailman set to send notifcations once a day).
I have been monitoring this list and the cpanel forums (http://forums.cpanel.net) to see if this issue is resolved. I have tried several of the fixes they had on there to no avail. Although they have set out a new fix http://forums.cpanel.net/showthread.php?s=&threadid=12740&highlight=mail man for the qrunner eating up cpu, but I haven't tried it yet. I am basically trying to find bits and pieces and make them into a whole that will fix my problems.
this is clipped and pasted from a final bounce message I am getting from MM. Since I don't really know how to read bounces I am not sure if the issue is on my side, their side or somewhere in the middle. Please note that the email that bounced on the sub notice is getting all its mail from just regular list mail.
Hope this helps! Back to the quest!
Chris
This is a Mailman mailing list bounce action notice:
List: Vegas
Member: WCDancer@aol.com
Action: Subscription disabled.
Reason: Excessive or fatal bounces.
The triggering bounce notice is attached below.
Questions? Contact the Mailman site administrator at mailman@domain.com
From: "Mail Delivery System" <Mailer-Daemon@alpha.domain1.com> To: mailman-bounces@alpha.domain2.com Subject: Mail delivery failed: returning message to sender Date: Mon, 28 Jul 2003 08:02:13 -0700
This message was created automatically by mail delivery software.
A message that you sent could not be delivered to one or more of its recipients. This is a permanent error. The following address(es) failed:
WCDancer@isp.com SMTP error from remote mailer after MAIL FROM:<mailman-bounces@alpha.domain2.com>: host mailin-01.mx.aol.com [152.163.224.26]: 550 REQUESTED ACTION NOT TAKEN: DNS FAILURE mcknz@isp.com SMTP error from remote mailer after MAIL FROM:<mailman-bounces@alpha.domain.com> SIZE=2643: host nymx-2.nyroc.rr.com [24.92.226.164]: 553 5.1.8 <mailman-bounces@alpha.domain.com>... Domain of sender address mailman-bounces@alpha.domain.com does not exist
------ This is a copy of the message, including all the headers.
Return-path: <mailman-bounces@alpha.domain2.com> Received: from alpha.domain2.com ([127.0.0.1]) by alpha.domain2.com with esmtp (Exim 4.20) id 19h9UO-0006jA-VB; Mon, 28 Jul 2003 08:00:40 -0700 Received: from [127.0.0.1] (helo=alpha.domain2.com) by alpha.domain.com with esmtp (Exim 4.20) id 19h9U7-0006h2-MD for Finewine-owner@bdomain2.com; Mon, 28 Jul 2003 08:00:23 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Subject: 3 Finewine moderator request(s) waiting From: Finewine-bounces@domain2.com To: Finewine-owner@domain2.com Message-ID: <mailman.11.1059404401.25640.finewine_domain2.com@domain2.com> Date: Mon, 28 Jul 2003 08:00:01 -0700 Precedence: bulk X-BeenThere: Finewine@domain.com X-Mailman-Version: 2.1.2 List-Id: X-List-Administrivia: yes Sender: mailman-bounces@alpha.domain.com Errors-To: mailman-bounces@alpha.domain.com
The Finewine@domain.com mailing list has 3 request(s) waiting for your consideration at:
Please attend to this at your earliest convenience. This notice of pending requests, if any, will be sent out daily.
Pending subscriptions: funnygirlster@isp.com (Sue) Sat Jul 19 17:23:45 2003 dpoday@isp.com (deborah) Thu Jul 24 22:38:08 2003 pupigrl1@isp.com (Jen) Sun Jul 27 04:36:20 2003
-----Original Message----- From: Richard Barrett [mailto:r.barrett@openinfo.co.uk] Sent: Thursday, July 31, 2003 7:30 PM To: Christine De La Rosa Cc: 'Mailman' Subject: Re: [Mailman-Users] Analysis of reasons for bounce responses
Chris
On Friday, August 1, 2003, at 12:48 am, Christine De La Rosa wrote:
This is the EXACT problem I am having with mailman, using cpanel and exim. I, however, am on a dedicated box and have root access. I have
As I said in my post, this bounce issue is not, in my view, at heart a
bug in Mailman. The evidence I saw of bounce notices from local MTA
(Exim) say that the MTA is having problems getting outgoing mail off
the system and finally bouncing it back to Mailman when it fails to do
so.
If you are getting final bounce notices from Mailman when it disables a
subscriber because of these problems what is the reason reported in the
bounce message from the MTA, which should be attached to the bottom of
the final bounce notification from MM? In the example final bounce
notice passed to me for comment the MTA was reporting it was having an
IP routing problem; that is hardly a Mailman generated problem as far
as I can tell.
One of my problems with looking at this is I am trying to help out a
fairly new user with a problem I have not seen on my systems, while I
have no access to the Mailman or MTA logs on a system experiencing the
problem. I have no standing with the CPanel owners to raise matters
with them in case they are contributing to the problem with the way
they set things up. Deeply frustrating.
Let me know if you gain any insight or information related to this.
I come from a majordomo background. When I got this dedicated server,
it had mailman pre-installed so I thought kewl I will use it instead.
Mailman can consume large amounts of CPU when the MTA it is trying to
hand mail over to repeatedly returns temporary failures. MM can be too
aggressive in repeatedly trying to get the MTA to either accept or
permanently fail the handover. There are changes in the Mailman CVS to
moderate MM's efforts in this respect but so far as I know those
changes are not available as a patch for MM 2.1.2 stable. This issue
may be affecting you but I am no expert myself in setting up Exim so as
to avoid this undesirable interaction; I do not see it with my Sendmail
setup. There have been posts to mailman-users on the topic of high CPU
utilisation. This one fromJon Carnes relates to Postfix MTA but its
substance/recommendations may be applicable to Exim:
http://www.mail-archive.com/mailman-users@python.org/msg17786.html
![](https://secure.gravatar.com/avatar/ab9b64c9ef8fe7aeb36f25bb1c5e5896.jpg?s=120&d=mm&r=g)
At 09:30 PM 7/31/2003, Christine De La Rosa wrote:
This is a big clue that the MTA is improperly configured and is trying to deliver to the AOL and Hotmail servers messages for more recipients than those systems allow - triggering anti-spam filtering on the recipient MTA at AOL and Hotmail. The messages are then bounced. My guess is that they changed the MTA at around the same time they upgraded Mailman, and you are barking up the wrong tree to try to find the source of the problem.
jc
![](https://secure.gravatar.com/avatar/a148cdd5c639fe49576e590c26f615ef.jpg?s=120&d=mm&r=g)
At 7:32 AM -0700 2003/08/02, JC Dill wrote:
A properly written MTA should automatically handle issues like
this and re-split the envelopes as necessary. Sendmail will do this, as will postfix. Dunno about Exim.
I think the problem lies somewhere with a locally screwed up MTA
which periodically won't respond at all, or can't get to its local DNS, so stuff bounces when it shouldn't. But I can't be sure.
-- Brad Knowles, <brad.knowles@skynet.be>
"They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -Benjamin Franklin, Historical Review of Pennsylvania.
GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E-(---) W+++(--) N+ !w--- O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++) tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)
![](https://secure.gravatar.com/avatar/e6041bf1a7b49ae3a76ff4859a1b03c7.jpg?s=120&d=mm&r=g)
On Sat, Aug 02, 2003 at 07:32:09AM -0700, JC Dill wrote:
A quick test would be to drop the maximum number of recipients in Mailman. In mm_cfg.py, set SMTP_MAX_RCPTS to 20. Details are in Defaults.py. This will chunk your messages to a maximum of 20 recipients at a time. Some people have done performance tests to suggest that dropping this to about 5 will give you better throughput.
-- Ed Wilts, Mounds View, MN, USA mailto:ewilts@ewilts.org
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
Thanks for the response JC. I am checkig with exim and cpanel, to see if they can at least point me in the right direction.
I checked my smtp-failure log in the mailman directory and are seeing hundreds of these.
Aug 03 08:01:14 2003 (30019) delivery to non_debutante@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:01:14 2003 (30019) delivery to queinelaiste@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:01:14 2003 (30019) delivery to wrangla@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:01:14 2003 (30019) delivery to fsteinnerd@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:03:22 2003 (30019) delivery to wolf@xxx.com failed with code 550: cannot route to sender address
Is this attached to the bounce backs I am getting which usually say 550 REQUESTED ACTION DNS Failure
Thanks for any help in advance
-----Original Message----- From: JC Dill [mailto:mailman@vo.cnchost.com] Sent: Saturday, August 02, 2003 9:32 AM To: Christine De La Rosa; 'Richard Barrett' Cc: 'Mailman' Subject: RE: [Mailman-Users] Analysis of reasons for bounce responses
At 09:30 PM 7/31/2003, Christine De La Rosa wrote:
Hi Richard,
Here is what I know so far, mail for most list members seem to be going
through for the most part. About two weeks ago MM mass unscribed about
This is a big clue that the MTA is improperly configured and is trying to deliver to the AOL and Hotmail servers messages for more recipients than
those systems allow - triggering anti-spam filtering on the recipient MTA at AOL and Hotmail. The messages are then bounced. My guess is that they changed the MTA at around the same time they upgraded Mailman, and you are barking up the wrong tree to try to find the source of the problem.
jc
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
Oh I forgot to add, that my dedicated is also a nameserver (domain.com) with about 10 hosted domains (domain1.com, domain2.com...)
Chris
-----Original Message----- From: Richard Barrett [mailto:r.barrett@openinfo.co.uk] Sent: Thursday, July 31, 2003 7:30 PM To: Christine De La Rosa Cc: 'Mailman' Subject: Re: [Mailman-Users] Analysis of reasons for bounce responses
Chris
On Friday, August 1, 2003, at 12:48 am, Christine De La Rosa wrote:
This is the EXACT problem I am having with mailman, using cpanel and exim. I, however, am on a dedicated box and have root access. I have
As I said in my post, this bounce issue is not, in my view, at heart a
bug in Mailman. The evidence I saw of bounce notices from local MTA
(Exim) say that the MTA is having problems getting outgoing mail off
the system and finally bouncing it back to Mailman when it fails to do
so.
If you are getting final bounce notices from Mailman when it disables a
subscriber because of these problems what is the reason reported in the
bounce message from the MTA, which should be attached to the bottom of
the final bounce notification from MM? In the example final bounce
notice passed to me for comment the MTA was reporting it was having an
IP routing problem; that is hardly a Mailman generated problem as far
as I can tell.
One of my problems with looking at this is I am trying to help out a
fairly new user with a problem I have not seen on my systems, while I
have no access to the Mailman or MTA logs on a system experiencing the
problem. I have no standing with the CPanel owners to raise matters
with them in case they are contributing to the problem with the way
they set things up. Deeply frustrating.
Let me know if you gain any insight or information related to this.
I come from a majordomo background. When I got this dedicated server,
it had mailman pre-installed so I thought kewl I will use it instead.
Mailman can consume large amounts of CPU when the MTA it is trying to
hand mail over to repeatedly returns temporary failures. MM can be too
aggressive in repeatedly trying to get the MTA to either accept or
permanently fail the handover. There are changes in the Mailman CVS to
moderate MM's efforts in this respect but so far as I know those
changes are not available as a patch for MM 2.1.2 stable. This issue
may be affecting you but I am no expert myself in setting up Exim so as
to avoid this undesirable interaction; I do not see it with my Sendmail
setup. There have been posts to mailman-users on the topic of high CPU
utilisation. This one fromJon Carnes relates to Postfix MTA but its
substance/recommendations may be applicable to Exim:
http://www.mail-archive.com/mailman-users@python.org/msg17786.html
![](https://secure.gravatar.com/avatar/d19f48f90665a9e11f86bc18f4836308.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Christine De La Rosa wrote:
Unfortunately, I work on a site which is hosted using cPanel. It's a far cry from the problem free experience I have with my own mailman installs which "just work." When they finally did the upgrade to 2.1.2, I had some immediate problems with bounces but then the server admin applied an update from cPanel that seems to have fixed that. I don't know it the bounces were at all similar to what you're having, but I mention it in case you haven't applied whatever updates cPanel has pushed out. It was either a mailman or exim update, I'm not sure since I don't admin the box.
Of course, the cPanel upgrade to 2.1.2 also broke some virtual host stuff for me that I have yet to get the host to work out. I have tried and tried to contact the folks at cPanel directly to find out what patches they apply to mailman and request the sources for those changes (as *required* by the GPL) to no avail. They have simply ignored me. I find that really annoying. If anyone knows of a way to reach someone at cPanel, please let us know. I'd like to point out to them that they *need* to make their changes available to the public. I'm not a fan of folks making money off of the work of open source software and then not even having the courtesy of honoring the few license requirements there are.
Todd OpenPGP -> KeyID: 0xD654075A | URL: www.pobox.com/~tmz/pgp
I got stopped by a cop the other day. He said, "Why'd you run that stop sign?" I said, "Because I don't believe everything I read." -- Stephen Wright
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3rc2 (GNU/Linux) Comment: When crypto is outlawed bayl bhgynjf jvyy unir cevinpl.
iD8DBQE/KbfEuv+09NZUB1oRAuC3AKCsj14RKywYjQtf3cXdIbFW2tOGtgCgsI7w jQdIuKN2/Ut9E4BwKyiOH/o= =27TD -----END PGP SIGNATURE-----
![](https://secure.gravatar.com/avatar/d4258e6caa5b231603e992b430199997.jpg?s=120&d=mm&r=g)
If anyone has sysadmin access to a server running CPanel, I would be happy do some trouble-shooting on the problem.
From the aggregate complaints it sounds like Cpanel has introduced an error with Exim. It could be something as simple as a timeout value.
Are the servers also running a local DNS server - and set to resolve using 127.0.0.1 as the primary nameserver? There are so many things that could be a problem...
On Thu, 2003-07-31 at 20:43, Todd wrote:
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
I checked my hosts file here it is
# Do not remove the following line, or various programs # that require network functionality will fail. 127.0.0.1 alpha.domain2.com localhost 64.62.144.106 alpha.domain2.com localhost
-----Original Message----- From: mailman-users-bounces+chris=christinedelarosa.com@python.org [mailto:mailman-users-bounces+chris=christinedelarosa.com@python.org] On Behalf Of Jon Carnes Sent: Thursday, July 31, 2003 10:47 PM To: Todd Cc: 'Mailman' Subject: Re: [Mailman-Users] Analysis of reasons for bounce responses
If anyone has sysadmin access to a server running CPanel, I would be happy do some trouble-shooting on the problem.
From the aggregate complaints it sounds like Cpanel has introduced an error with Exim. It could be something as simple as a timeout value.
Are the servers also running a local DNS server - and set to resolve using 127.0.0.1 as the primary nameserver? There are so many things that could be a problem...
On Thu, 2003-07-31 at 20:43, Todd wrote:
Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
This message was sent to: chris@christinedelarosa.com Unsubscribe or change your options at http://mail.python.org/mailman/options/mailman-users/chris%40christinede larosa.com
![](https://secure.gravatar.com/avatar/d4258e6caa5b231603e992b430199997.jpg?s=120&d=mm&r=g)
On Thu, 2003-07-31 at 06:08, Richard Barrett wrote:
I have seen this behavior a few times and in each case it was resource problem on the server. They need to check on the memory of the box and how much is being used during the bounces. I'm afraid it is only something that the sysadmin can really do.
I have also seen this problem when the /var directory was made too small. Could this be a problem with the users services being chrooted and them running out of quota space while processing large messages?
Good Luck - Jon Carnes
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
This is the EXACT problem I am having with mailman, using cpanel and exim. I, however, am on a dedicated box and have root access. I have been working on this problem for over three weeks. I am tired and frustrated and wishing I knew a little more about mailman and its inner workings. I come from a majordomo background. When I got this dedicated server, it had mailman pre-installed so I thought kewl I will use it instead. It ran fabulously until cpanel did their last upgrade. I have been in hell since then.
I wish I could help you, but I have not been able to figure out even from the logs what the issues are. Granted, looking at the logs are like gobbledigook to me. So I have spent a lot of time searching the archives for this problem. Although, I haven't been able to find anyone with my specific issue. I can tell you that qrunner has been consuming my cpu at astonishing rates. But there is no rhyme or reason to that, that I can see. If I find anything out I will post it here.
Chris
-----Original Message----- From: mailman-users-bounces+chris=christinedelarosa.com@python.org [mailto:mailman-users-bounces+chris=christinedelarosa.com@python.org] On Behalf Of Richard Barrett Sent: Thursday, July 31, 2003 5:08 AM To: Mailman Subject: [Mailman-Users] Analysis of reasons for bounce responses
This is not my problem but one soemone approached me about off list.
I'll summarise their situation and I would appreciate any thoughts,
comments, sanity checking on my analysis.
Situation:
List admin's domain is hosted by a server run by a hosting company.
The server is (probably) running Linux
Mailman 2.1.2 is provided on the server with, I believe, CPanel
The server's MTA appears to be Exim 4.20 handling both incoming and
outgoing messages to/from MailmanMailman list's basically work OK much of the time
Periodically messages out from Mailman are being bounced en masse
leading to the subscriber accounts being disabledThe bounce response attached to the bounce action notice produced by
Mailman when subscribers are disabled in these circumstances is from
the server local MTA. The MTA lists what appears to be all the
addresses to which the outgoing message was sent with a note against
each saying "unrouteable mail domain maildomain.tld". The bounce
response asserts that this is a permanent failure. The messages are
each being sent to a number of different mail domains not just a single
one
The list admin with the problem has no access to either the Mailman
or MTA logsThe server administrator is supposedly claiming that these bounces
are a result of "a bug in Mailman for which no patch is available"
My analysis is:
The problem is not obviously due to any bug in Mailman
The problem is more likely due to an IP routing problem which
periodically afflicts the server and its ability to reach the MTAs for
the mail domains addressed by the outgoing messagesThe MTA is accepting the outgoing messages from Mailman and then
finding it has problems of its own in delivering mail which are
unrelated to Mailman as the origin of the messages
But I have to concede I know nothing about Exim 4.20. I know even less
about CPanel.
Any thoughts or anatomically feasible suggestions welcome.
Richard Barrett
http://www.openinfo.co.uk
Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
This message was sent to: chris@christinedelarosa.com Unsubscribe or change your options at http://mail.python.org/mailman/options/mailman-users/chris%40christinede larosa.com
![](https://secure.gravatar.com/avatar/632bc87a86b0db6153c56defac0ba682.jpg?s=120&d=mm&r=g)
Chris
On Friday, August 1, 2003, at 12:48 am, Christine De La Rosa wrote:
As I said in my post, this bounce issue is not, in my view, at heart a
bug in Mailman. The evidence I saw of bounce notices from local MTA
(Exim) say that the MTA is having problems getting outgoing mail off
the system and finally bouncing it back to Mailman when it fails to do
so.
If you are getting final bounce notices from Mailman when it disables a
subscriber because of these problems what is the reason reported in the
bounce message from the MTA, which should be attached to the bottom of
the final bounce notification from MM? In the example final bounce
notice passed to me for comment the MTA was reporting it was having an
IP routing problem; that is hardly a Mailman generated problem as far
as I can tell.
One of my problems with looking at this is I am trying to help out a
fairly new user with a problem I have not seen on my systems, while I
have no access to the Mailman or MTA logs on a system experiencing the
problem. I have no standing with the CPanel owners to raise matters
with them in case they are contributing to the problem with the way
they set things up. Deeply frustrating.
Let me know if you gain any insight or information related to this.
Mailman can consume large amounts of CPU when the MTA it is trying to
hand mail over to repeatedly returns temporary failures. MM can be too
aggressive in repeatedly trying to get the MTA to either accept or
permanently fail the handover. There are changes in the Mailman CVS to
moderate MM's efforts in this respect but so far as I know those
changes are not available as a patch for MM 2.1.2 stable. This issue
may be affecting you but I am no expert myself in setting up Exim so as
to avoid this undesirable interaction; I do not see it with my Sendmail
setup. There have been posts to mailman-users on the topic of high CPU
utilisation. This one fromJon Carnes relates to Postfix MTA but its
substance/recommendations may be applicable to Exim:
http://www.mail-archive.com/mailman-users@python.org/msg17786.html
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
Hi Richard,
Here is what I know so far, mail for most list members seem to be going through for the most part. About two weeks ago MM mass unscribed about 300 folx. Now we seem to have stablized the members but the owners of the lists are getting routinely bounced. Yahoo email address do not have a problem, aol is a major headache as is hotmail. What is happening on my side is that the sub notices are never reaching the owners, so they are bouncing like mad and eventually hit the excessive bounce phase after five days (since we have the settings set to set to nomail after 5 bounces and mailman set to send notifcations once a day).
I have been monitoring this list and the cpanel forums (http://forums.cpanel.net) to see if this issue is resolved. I have tried several of the fixes they had on there to no avail. Although they have set out a new fix http://forums.cpanel.net/showthread.php?s=&threadid=12740&highlight=mail man for the qrunner eating up cpu, but I haven't tried it yet. I am basically trying to find bits and pieces and make them into a whole that will fix my problems.
this is clipped and pasted from a final bounce message I am getting from MM. Since I don't really know how to read bounces I am not sure if the issue is on my side, their side or somewhere in the middle. Please note that the email that bounced on the sub notice is getting all its mail from just regular list mail.
Hope this helps! Back to the quest!
Chris
This is a Mailman mailing list bounce action notice:
List: Vegas
Member: WCDancer@aol.com
Action: Subscription disabled.
Reason: Excessive or fatal bounces.
The triggering bounce notice is attached below.
Questions? Contact the Mailman site administrator at mailman@domain.com
From: "Mail Delivery System" <Mailer-Daemon@alpha.domain1.com> To: mailman-bounces@alpha.domain2.com Subject: Mail delivery failed: returning message to sender Date: Mon, 28 Jul 2003 08:02:13 -0700
This message was created automatically by mail delivery software.
A message that you sent could not be delivered to one or more of its recipients. This is a permanent error. The following address(es) failed:
WCDancer@isp.com SMTP error from remote mailer after MAIL FROM:<mailman-bounces@alpha.domain2.com>: host mailin-01.mx.aol.com [152.163.224.26]: 550 REQUESTED ACTION NOT TAKEN: DNS FAILURE mcknz@isp.com SMTP error from remote mailer after MAIL FROM:<mailman-bounces@alpha.domain.com> SIZE=2643: host nymx-2.nyroc.rr.com [24.92.226.164]: 553 5.1.8 <mailman-bounces@alpha.domain.com>... Domain of sender address mailman-bounces@alpha.domain.com does not exist
------ This is a copy of the message, including all the headers.
Return-path: <mailman-bounces@alpha.domain2.com> Received: from alpha.domain2.com ([127.0.0.1]) by alpha.domain2.com with esmtp (Exim 4.20) id 19h9UO-0006jA-VB; Mon, 28 Jul 2003 08:00:40 -0700 Received: from [127.0.0.1] (helo=alpha.domain2.com) by alpha.domain.com with esmtp (Exim 4.20) id 19h9U7-0006h2-MD for Finewine-owner@bdomain2.com; Mon, 28 Jul 2003 08:00:23 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Subject: 3 Finewine moderator request(s) waiting From: Finewine-bounces@domain2.com To: Finewine-owner@domain2.com Message-ID: <mailman.11.1059404401.25640.finewine_domain2.com@domain2.com> Date: Mon, 28 Jul 2003 08:00:01 -0700 Precedence: bulk X-BeenThere: Finewine@domain.com X-Mailman-Version: 2.1.2 List-Id: X-List-Administrivia: yes Sender: mailman-bounces@alpha.domain.com Errors-To: mailman-bounces@alpha.domain.com
The Finewine@domain.com mailing list has 3 request(s) waiting for your consideration at:
Please attend to this at your earliest convenience. This notice of pending requests, if any, will be sent out daily.
Pending subscriptions: funnygirlster@isp.com (Sue) Sat Jul 19 17:23:45 2003 dpoday@isp.com (deborah) Thu Jul 24 22:38:08 2003 pupigrl1@isp.com (Jen) Sun Jul 27 04:36:20 2003
-----Original Message----- From: Richard Barrett [mailto:r.barrett@openinfo.co.uk] Sent: Thursday, July 31, 2003 7:30 PM To: Christine De La Rosa Cc: 'Mailman' Subject: Re: [Mailman-Users] Analysis of reasons for bounce responses
Chris
On Friday, August 1, 2003, at 12:48 am, Christine De La Rosa wrote:
This is the EXACT problem I am having with mailman, using cpanel and exim. I, however, am on a dedicated box and have root access. I have
As I said in my post, this bounce issue is not, in my view, at heart a
bug in Mailman. The evidence I saw of bounce notices from local MTA
(Exim) say that the MTA is having problems getting outgoing mail off
the system and finally bouncing it back to Mailman when it fails to do
so.
If you are getting final bounce notices from Mailman when it disables a
subscriber because of these problems what is the reason reported in the
bounce message from the MTA, which should be attached to the bottom of
the final bounce notification from MM? In the example final bounce
notice passed to me for comment the MTA was reporting it was having an
IP routing problem; that is hardly a Mailman generated problem as far
as I can tell.
One of my problems with looking at this is I am trying to help out a
fairly new user with a problem I have not seen on my systems, while I
have no access to the Mailman or MTA logs on a system experiencing the
problem. I have no standing with the CPanel owners to raise matters
with them in case they are contributing to the problem with the way
they set things up. Deeply frustrating.
Let me know if you gain any insight or information related to this.
I come from a majordomo background. When I got this dedicated server,
it had mailman pre-installed so I thought kewl I will use it instead.
Mailman can consume large amounts of CPU when the MTA it is trying to
hand mail over to repeatedly returns temporary failures. MM can be too
aggressive in repeatedly trying to get the MTA to either accept or
permanently fail the handover. There are changes in the Mailman CVS to
moderate MM's efforts in this respect but so far as I know those
changes are not available as a patch for MM 2.1.2 stable. This issue
may be affecting you but I am no expert myself in setting up Exim so as
to avoid this undesirable interaction; I do not see it with my Sendmail
setup. There have been posts to mailman-users on the topic of high CPU
utilisation. This one fromJon Carnes relates to Postfix MTA but its
substance/recommendations may be applicable to Exim:
http://www.mail-archive.com/mailman-users@python.org/msg17786.html
![](https://secure.gravatar.com/avatar/ab9b64c9ef8fe7aeb36f25bb1c5e5896.jpg?s=120&d=mm&r=g)
At 09:30 PM 7/31/2003, Christine De La Rosa wrote:
This is a big clue that the MTA is improperly configured and is trying to deliver to the AOL and Hotmail servers messages for more recipients than those systems allow - triggering anti-spam filtering on the recipient MTA at AOL and Hotmail. The messages are then bounced. My guess is that they changed the MTA at around the same time they upgraded Mailman, and you are barking up the wrong tree to try to find the source of the problem.
jc
![](https://secure.gravatar.com/avatar/a148cdd5c639fe49576e590c26f615ef.jpg?s=120&d=mm&r=g)
At 7:32 AM -0700 2003/08/02, JC Dill wrote:
A properly written MTA should automatically handle issues like
this and re-split the envelopes as necessary. Sendmail will do this, as will postfix. Dunno about Exim.
I think the problem lies somewhere with a locally screwed up MTA
which periodically won't respond at all, or can't get to its local DNS, so stuff bounces when it shouldn't. But I can't be sure.
-- Brad Knowles, <brad.knowles@skynet.be>
"They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -Benjamin Franklin, Historical Review of Pennsylvania.
GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E-(---) W+++(--) N+ !w--- O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++) tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)
![](https://secure.gravatar.com/avatar/e6041bf1a7b49ae3a76ff4859a1b03c7.jpg?s=120&d=mm&r=g)
On Sat, Aug 02, 2003 at 07:32:09AM -0700, JC Dill wrote:
A quick test would be to drop the maximum number of recipients in Mailman. In mm_cfg.py, set SMTP_MAX_RCPTS to 20. Details are in Defaults.py. This will chunk your messages to a maximum of 20 recipients at a time. Some people have done performance tests to suggest that dropping this to about 5 will give you better throughput.
-- Ed Wilts, Mounds View, MN, USA mailto:ewilts@ewilts.org
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
Thanks for the response JC. I am checkig with exim and cpanel, to see if they can at least point me in the right direction.
I checked my smtp-failure log in the mailman directory and are seeing hundreds of these.
Aug 03 08:01:14 2003 (30019) delivery to non_debutante@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:01:14 2003 (30019) delivery to queinelaiste@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:01:14 2003 (30019) delivery to wrangla@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:01:14 2003 (30019) delivery to fsteinnerd@xxx.com failed with code 550: cannot route to sender address <mailman-bounces@alpha.butch-femme.com> Aug 03 08:03:22 2003 (30019) delivery to wolf@xxx.com failed with code 550: cannot route to sender address
Is this attached to the bounce backs I am getting which usually say 550 REQUESTED ACTION DNS Failure
Thanks for any help in advance
-----Original Message----- From: JC Dill [mailto:mailman@vo.cnchost.com] Sent: Saturday, August 02, 2003 9:32 AM To: Christine De La Rosa; 'Richard Barrett' Cc: 'Mailman' Subject: RE: [Mailman-Users] Analysis of reasons for bounce responses
At 09:30 PM 7/31/2003, Christine De La Rosa wrote:
Hi Richard,
Here is what I know so far, mail for most list members seem to be going
through for the most part. About two weeks ago MM mass unscribed about
This is a big clue that the MTA is improperly configured and is trying to deliver to the AOL and Hotmail servers messages for more recipients than
those systems allow - triggering anti-spam filtering on the recipient MTA at AOL and Hotmail. The messages are then bounced. My guess is that they changed the MTA at around the same time they upgraded Mailman, and you are barking up the wrong tree to try to find the source of the problem.
jc
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
Oh I forgot to add, that my dedicated is also a nameserver (domain.com) with about 10 hosted domains (domain1.com, domain2.com...)
Chris
-----Original Message----- From: Richard Barrett [mailto:r.barrett@openinfo.co.uk] Sent: Thursday, July 31, 2003 7:30 PM To: Christine De La Rosa Cc: 'Mailman' Subject: Re: [Mailman-Users] Analysis of reasons for bounce responses
Chris
On Friday, August 1, 2003, at 12:48 am, Christine De La Rosa wrote:
This is the EXACT problem I am having with mailman, using cpanel and exim. I, however, am on a dedicated box and have root access. I have
As I said in my post, this bounce issue is not, in my view, at heart a
bug in Mailman. The evidence I saw of bounce notices from local MTA
(Exim) say that the MTA is having problems getting outgoing mail off
the system and finally bouncing it back to Mailman when it fails to do
so.
If you are getting final bounce notices from Mailman when it disables a
subscriber because of these problems what is the reason reported in the
bounce message from the MTA, which should be attached to the bottom of
the final bounce notification from MM? In the example final bounce
notice passed to me for comment the MTA was reporting it was having an
IP routing problem; that is hardly a Mailman generated problem as far
as I can tell.
One of my problems with looking at this is I am trying to help out a
fairly new user with a problem I have not seen on my systems, while I
have no access to the Mailman or MTA logs on a system experiencing the
problem. I have no standing with the CPanel owners to raise matters
with them in case they are contributing to the problem with the way
they set things up. Deeply frustrating.
Let me know if you gain any insight or information related to this.
I come from a majordomo background. When I got this dedicated server,
it had mailman pre-installed so I thought kewl I will use it instead.
Mailman can consume large amounts of CPU when the MTA it is trying to
hand mail over to repeatedly returns temporary failures. MM can be too
aggressive in repeatedly trying to get the MTA to either accept or
permanently fail the handover. There are changes in the Mailman CVS to
moderate MM's efforts in this respect but so far as I know those
changes are not available as a patch for MM 2.1.2 stable. This issue
may be affecting you but I am no expert myself in setting up Exim so as
to avoid this undesirable interaction; I do not see it with my Sendmail
setup. There have been posts to mailman-users on the topic of high CPU
utilisation. This one fromJon Carnes relates to Postfix MTA but its
substance/recommendations may be applicable to Exim:
http://www.mail-archive.com/mailman-users@python.org/msg17786.html
![](https://secure.gravatar.com/avatar/d19f48f90665a9e11f86bc18f4836308.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Christine De La Rosa wrote:
Unfortunately, I work on a site which is hosted using cPanel. It's a far cry from the problem free experience I have with my own mailman installs which "just work." When they finally did the upgrade to 2.1.2, I had some immediate problems with bounces but then the server admin applied an update from cPanel that seems to have fixed that. I don't know it the bounces were at all similar to what you're having, but I mention it in case you haven't applied whatever updates cPanel has pushed out. It was either a mailman or exim update, I'm not sure since I don't admin the box.
Of course, the cPanel upgrade to 2.1.2 also broke some virtual host stuff for me that I have yet to get the host to work out. I have tried and tried to contact the folks at cPanel directly to find out what patches they apply to mailman and request the sources for those changes (as *required* by the GPL) to no avail. They have simply ignored me. I find that really annoying. If anyone knows of a way to reach someone at cPanel, please let us know. I'd like to point out to them that they *need* to make their changes available to the public. I'm not a fan of folks making money off of the work of open source software and then not even having the courtesy of honoring the few license requirements there are.
Todd OpenPGP -> KeyID: 0xD654075A | URL: www.pobox.com/~tmz/pgp
I got stopped by a cop the other day. He said, "Why'd you run that stop sign?" I said, "Because I don't believe everything I read." -- Stephen Wright
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3rc2 (GNU/Linux) Comment: When crypto is outlawed bayl bhgynjf jvyy unir cevinpl.
iD8DBQE/KbfEuv+09NZUB1oRAuC3AKCsj14RKywYjQtf3cXdIbFW2tOGtgCgsI7w jQdIuKN2/Ut9E4BwKyiOH/o= =27TD -----END PGP SIGNATURE-----
![](https://secure.gravatar.com/avatar/d4258e6caa5b231603e992b430199997.jpg?s=120&d=mm&r=g)
If anyone has sysadmin access to a server running CPanel, I would be happy do some trouble-shooting on the problem.
From the aggregate complaints it sounds like Cpanel has introduced an error with Exim. It could be something as simple as a timeout value.
Are the servers also running a local DNS server - and set to resolve using 127.0.0.1 as the primary nameserver? There are so many things that could be a problem...
On Thu, 2003-07-31 at 20:43, Todd wrote:
![](https://secure.gravatar.com/avatar/7d8d8a1e1bafd44add3e8dad15ddcfd3.jpg?s=120&d=mm&r=g)
I checked my hosts file here it is
# Do not remove the following line, or various programs # that require network functionality will fail. 127.0.0.1 alpha.domain2.com localhost 64.62.144.106 alpha.domain2.com localhost
-----Original Message----- From: mailman-users-bounces+chris=christinedelarosa.com@python.org [mailto:mailman-users-bounces+chris=christinedelarosa.com@python.org] On Behalf Of Jon Carnes Sent: Thursday, July 31, 2003 10:47 PM To: Todd Cc: 'Mailman' Subject: Re: [Mailman-Users] Analysis of reasons for bounce responses
If anyone has sysadmin access to a server running CPanel, I would be happy do some trouble-shooting on the problem.
From the aggregate complaints it sounds like Cpanel has introduced an error with Exim. It could be something as simple as a timeout value.
Are the servers also running a local DNS server - and set to resolve using 127.0.0.1 as the primary nameserver? There are so many things that could be a problem...
On Thu, 2003-07-31 at 20:43, Todd wrote:
Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
This message was sent to: chris@christinedelarosa.com Unsubscribe or change your options at http://mail.python.org/mailman/options/mailman-users/chris%40christinede larosa.com
participants (7)
-
Brad Knowles
-
Christine De La Rosa
-
Ed Wilts
-
JC Dill
-
Jon Carnes
-
Richard Barrett
-
Todd