The qpopper list archive ending on 18 Nov 2005
Topics covered in this issue include:
1. Re: qpopper 4.0.8 + mysql
Daniel Senie <dts at senie dot com>
Fri, 12 Aug 2005 19:47:27 -0400
2. Re: qpopper 4.0.8 + mysql
Martin Kellermann <Kellermann at sk-datentechnik dot com>
Mon, 15 Aug 2005 13:40:19 +0200
3. Re: qpopper 4.0.8 + mysql
Martin Kellermann <Kellermann at sk-datentechnik dot com>
Mon, 15 Aug 2005 15:48:40 +0200
4. Re: qpopper 4.0.8 + mysql
Ken A <ka at pacific dot net>
Mon, 15 Aug 2005 08:42:56 -0700
5. Re: qpopper 4.0.8 + mysql
Ken A <ka at pacific dot net>
Mon, 15 Aug 2005 08:59:56 -0700
6. Re: qpopper 4.0.8 + mysql
Martin Kellermann <Kellermann at sk-datentechnik dot com>
Wed, 17 Aug 2005 17:25:50 +0200
7. empty lines in logfile
"M dot Kellermann" <Kellermann at sk-datentechnik dot com>
Thu, 08 Sep 2005 13:36:49 +0200
8. Re: empty lines in logfile
Daniel Senie <dts at senie dot com>
Thu, 08 Sep 2005 10:21:56 -0400
9. Re: empty lines in logfile
"M dot Kellermann" <kellermann at sk-datentechnik dot com>
Thu, 08 Sep 2005 18:07:23 +0200
10. Re: empty lines in logfile
Daniel Senie <dts at senie dot com>
Thu, 08 Sep 2005 12:58:20 -0400
11. Re: empty lines in logfile
Randall Gellens <randy at qualcomm dot com>
Thu, 8 Sep 2005 16:06:24 -0700
12. Re: empty lines in logfile
"M dot Kellermann" <kellermann at sk-datentechnik dot com>
Fri, 09 Sep 2005 08:50:19 +0200
13. [solved] Re: empty lines in logfile
"M dot Kellermann" <kellermann at sk-datentechnik dot com>
Fri, 09 Sep 2005 09:04:20 +0200
14. qpopper ***glibc detected *** malloc() : memory corruption
Ken A <ka at pacific dot net>
Mon, 12 Sep 2005 08:18:16 -0700
15. re: qpopper ***glibc detected *** malloc() : memory corruption
Ken A <ka at pacific dot net>
Mon, 12 Sep 2005 08:29:52 -0700
16. re: qpopper ***glibc detected *** malloc() : memory corruption
Randall Gellens <randy at qualcomm dot com>
Wed, 14 Sep 2005 10:17:17 -0700
17. Re: qpopper ***glibc detected *** malloc() : memory corruption
Ken A <ka at pacific dot net>
Wed, 14 Sep 2005 11:01:14 -0700
18. error flushing output to client
"GioiaBa" <gioiaba at slacky dot it>
Mon, 19 Sep 2005 14:56:30 +0200
19. Re: error flushing output to client
Daniel Senie <dts at senie dot com>
Mon, 19 Sep 2005 09:33:00 -0400
20. Re: error flushing output to client
"GioiaBa" <gioiaba at slacky dot it>
Mon, 19 Sep 2005 16:41:45 +0200
21. Re: error flushing output to client
Alan Brown <alanb at digistar dot com>
Mon, 19 Sep 2005 12:19:26 -0400 (EDT)
22. "Password supplied is incorrect" error with Solaris 9...
"Doug Wellington" <ddw at rinconresearch dot com>
Fri, 23 Sep 2005 16:31:33 -0700
23. Re: ***spam*** "Password supplied is incorrect" error with Solaris 9...
Clifton Royston <cliftonr at lava dot net>
Fri, 23 Sep 2005 15:20:08 -1000
24. Re: "Password supplied is incorrect" error with Solaris 9...
Jerry K <qpopper at oryx dot cc>
Mon, 26 Sep 2005 12:48:45 -0500
25. qpopper and quotas
"Edward Chase" <echase at studentweb dot providence dot edu>
Mon, 26 Sep 2005 15:04:09 -0400
26. Re: qpopper and quotas
Daniel Senie <dts at senie dot com>
Mon, 26 Sep 2005 15:31:23 -0400
27. Re: qpopper and quotas
Ken A <ka at pacific dot net>
Mon, 26 Sep 2005 12:28:46 -0700
28. Re: qpopper and quotas
Clifton Royston <cliftonr at lava dot net>
Mon, 26 Sep 2005 10:02:54 -1000
29. Re: qpopper and quotas
Ken A <ka at pacific dot net>
Mon, 26 Sep 2005 14:44:28 -0700
30. Re: qpopper and quotas
David Champion <dgc at uchicago dot edu>
Mon, 26 Sep 2005 15:35:36 -0500
31. Re: qpopper and quotas
Alan Brown <alanb at digistar dot com>
Mon, 26 Sep 2005 17:22:20 -0400 (EDT)
32. RE: qpopper and quotas
"Alan W dot Rateliff, II" <lists at rateliff dot net>
Tue, 27 Sep 2005 10:49:04 -0400
33. RE: qpopper and quotas
"Edward Chase" <echase at studentweb dot providence dot edu>
Wed, 28 Sep 2005 07:52:02 -0400
34. RE: qpopper and quotas
"Edward Chase" <echase at studentweb dot providence dot edu>
Wed, 28 Sep 2005 07:52:02 -0400
35. Re: qpopper and quotas
Spiros Ioannou <sivann at image dot ece dot ntua dot gr>
Wed, 28 Sep 2005 15:26:21 +0300
36. RE: qpopper and quotas
"Edward Chase" <echase at studentweb dot providence dot edu>
Wed, 28 Sep 2005 08:13:39 -0400
37. Re: qpopper and quotas
Clifton Royston <cliftonr at lava dot net>
Wed, 28 Sep 2005 08:47:40 -1000
38. Re: qpopper and quotas
Ken A <ka at pacific dot net>
Wed, 28 Sep 2005 14:14:39 -0700
39. RE: qpopper and quotas
Alan Brown <alanb at digistar dot com>
Wed, 28 Sep 2005 15:17:13 -0400 (EDT)
40. Re: qpopper and quotas - a solution
Spiros Ioannou <sivann at image dot ece dot ntua dot gr>
Thu, 29 Sep 2005 14:05:20 +0300
41. OT - imapd (was qpopper and quotas)
Edward Chase <echase at studentweb dot providence dot edu>
Thu, 29 Sep 2005 18:06:47 -0400 (EDT)
42. Re: OT - imapd (was qpopper and quotas)
Clifton Royston <cliftonr at lava dot net>
Thu, 29 Sep 2005 13:52:28 -1000
43. Re: qpopper and quotas - a solution
Clifton Royston <cliftonr at lava dot net>
Tue, 4 Oct 2005 07:06:44 -1000
44. Re: qpopper and quotas - a solution
Spiros Ioannou <sivann at image dot ece dot ntua dot gr>
Mon, 03 Oct 2005 17:11:02 +0300
45. Broken pipe and timeout errors
Kai Lanz <lanz at pangea dot Stanford dot EDU>
Thu, 13 Oct 2005 17:38:24 -0700 (PDT)
46. RE: Broken pipe and timeout errors
"Edward Chase" <echase at studentweb dot providence dot edu>
Fri, 14 Oct 2005 12:53:17 -0400
47. RE: Broken pipe and timeout errors
Alan Brown <alanb at digistar dot com>
Fri, 14 Oct 2005 14:31:42 -0400 (EDT)
48. Re: "Password supplied is incorrect" error with Solaris 9...
Randall Gellens <randy at qualcomm dot com>
Sun, 16 Oct 2005 23:25:04 -0700
49. getting poppassd on fedora 3 to work?
Tim Tyler <tyler at beloit dot edu>
Fri, 18 Nov 2005 14:55:46 -0600
50. Re: getting poppassd on fedora 3 to work?
Daniel Senie <dts at senie dot com>
Fri, 18 Nov 2005 16:09:19 -0500
Date: Fri, 12 Aug 2005 19:47:27 -0400
From: Daniel Senie <dts at senie dot com>
Subject: Re: qpopper 4.0.8 + mysql
At 06:36 PM 8/12/2005, Ken A wrote:
>An alternative to patching qpopper is pam_mysql, which works fine
>with 4.08. Just configure qpopper --with-pam and setup
>/etc/pam.d/pop3 to use pam_mysql.so. This makes upgrades a bit less painful.
I like the sound of that!
Perhaps all we need at this point is some text for the documentation,
rather than software work :)
>Dan
>
>Martin Kellermann wrote:
>>hi list,
>>is there a patch for qpopper 4.0.8 to support mysql auth?
>>what i found is:
>>http://www.asteroid-b612.org/software/qpopper-mysql/qpopper-mysql-0.14.patch
>>but unfortunately it is for version 4.0.5 ...
>>or, does it work for 4.0.8 ?
>>i use qpopper for lots of years now....simply wonderful.
>>but some mysql support would be fine...
>>thanks for you hints..
>>MK
Date: Mon, 15 Aug 2005 13:40:19 +0200
From: Martin Kellermann <Kellermann at sk-datentechnik dot com>
Subject: Re: qpopper 4.0.8 + mysql
thanks for the hint, i will try pam_mysql the next days...
another question:
is there a known way/patch to get qpopper 4.0.8 work with the cyrus-sasl2 lib?
thank you
MK
At 01:47 13.08.2005, you wrote:
>At 06:36 PM 8/12/2005, Ken A wrote:
>>An alternative to patching qpopper is pam_mysql, which works fine
>>with 4.08. Just configure qpopper --with-pam and setup
>>/etc/pam.d/pop3 to use pam_mysql.so. This makes upgrades a bit less painful.
>
>I like the sound of that!
>
>Perhaps all we need at this point is some text for the
>documentation, rather than software work :)
>
>>Dan
>>
>>Martin Kellermann wrote:
>>>hi list,
>>>is there a patch for qpopper 4.0.8 to support mysql auth?
>>>what i found is:
>>>http://www.asteroid-b612.org/software/qpopper-mysql/qpopper-mysql-0.14.patch
>>>but unfortunately it is for version 4.0.5 ...
>>>or, does it work for 4.0.8 ?
>>>i use qpopper for lots of years now....simply wonderful.
>>>but some mysql support would be fine...
>>>thanks for you hints..
>>>MK
Date: Mon, 15 Aug 2005 15:48:40 +0200
From: Martin Kellermann <Kellermann at sk-datentechnik dot com>
Subject: Re: qpopper 4.0.8 + mysql
hi Ken,
i was too curious and tried qpopper/PAM/mysql.
it is up and running... thanks!
but there is one strange behaviour which i cant explain...
users are ONLY authentivated if they have a local account (etc/passwd)!?
so, maybe my question is a bit offtopic, but i think it is maybe a
qpopper or config problem.
case 1:
- local user named "test" has local password "test"
- mysql database has username "test" and password "otherthantest"
- qpopper/pam authenticates the user correct when logging in with
user "test" and pass "otherthantest"
- pam_mysql syslogs without errors, everything is fine
case 2:
- no local user test exists
- mysql database has username "test" and password "otherthantest"
- qpopper/pam DOES NOT authenticate the user logging in with user
"test" and pass "otherthantest"
- pam_mysql syslogs shows NOTHING !
it appears that, when there is no corresponding local user, the whole
PAM mechanism isnt started...
i configured qpopper with:
./configure --prefix=/usr --sysconfdir=/etc --enable-shy --with-pam=pop3
my /etc/pam.conf is this:
pop3 auth required pam_mysql.so verbose=1 host=localhost db=* table=*
user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
pop3 account required pam_mysql.so verbose=1 host=localhost db=*
table=* user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
any ideas?
is there a way to tell qpopper to ONLY use PAM for authentication?
thanks in advance
MK
At 00:36 13.08.2005, you wrote:
>An alternative to patching qpopper is pam_mysql, which works fine
>with 4.08. Just configure qpopper --with-pam and setup
>/etc/pam.d/pop3 to use pam_mysql.so. This makes upgrades a bit less painful.
>
>Ken
>
>
>Martin Kellermann wrote:
>>hi list,
>>is there a patch for qpopper 4.0.8 to support mysql auth?
>>what i found is:
>>http://www.asteroid-b612.org/software/qpopper-mysql/qpopper-mysql-0.14.patch
>>but unfortunately it is for version 4.0.5 ...
>>or, does it work for 4.0.8 ?
>>i use qpopper for lots of years now....simply wonderful.
>>but some mysql support would be fine...
>>thanks for you hints..
>>MK
Date: Mon, 15 Aug 2005 08:42:56 -0700
From: Ken A <ka at pacific dot net>
Subject: Re: qpopper 4.0.8 + mysql
Qpopper changes uids to read/write spools and temp drop files.
Authentication is just one part of PAM. AFAIK, pam_mysql doesn't yet
implement PAM 'account' functions, so you must have a local uid in
/etc/passwd. The local user need not have a password. I simply create
the local users with no home dir, no shell, no password, etc.
Another possibility is using NSS to put uids in mysql, and PAM together,
but I have not tried this. Here's an interesting doc that explains some
of these issues.
http://users.linuxbourg.ch/fvgoto/informatica/tbsc/doc/final/pam_mysqlim.pdf
Ken A.
Martin Kellermann wrote:
> hi Ken,
>
> i was too curious and tried qpopper/PAM/mysql.
> it is up and running... thanks!
>
> but there is one strange behaviour which i cant explain...
> users are ONLY authentivated if they have a local account (etc/passwd)!?
>
> so, maybe my question is a bit offtopic, but i think it is maybe a
> qpopper or config problem.
>
> case 1:
> - local user named "test" has local password "test"
> - mysql database has username "test" and password "otherthantest"
> - qpopper/pam authenticates the user correct when logging in with user
> "test" and pass "otherthantest"
> - pam_mysql syslogs without errors, everything is fine
>
> case 2:
> - no local user test exists
> - mysql database has username "test" and password "otherthantest"
> - qpopper/pam DOES NOT authenticate the user logging in with user "test"
> and pass "otherthantest"
> - pam_mysql syslogs shows NOTHING !
> it appears that, when there is no corresponding local user, the whole
> PAM mechanism isnt started...
>
> i configured qpopper with:
> ./configure --prefix=/usr --sysconfdir=/etc --enable-shy --with-pam=pop3
>
> my /etc/pam.conf is this:
> pop3 auth required pam_mysql.so verbose=1 host=localhost db=* table=*
> user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
> pop3 account required pam_mysql.so verbose=1 host=localhost db=* table=*
> user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
>
> any ideas?
> is there a way to tell qpopper to ONLY use PAM for authentication?
>
> thanks in advance
>
> MK
>
> At 00:36 13.08.2005, you wrote:
>> An alternative to patching qpopper is pam_mysql, which works fine with
>> 4.08. Just configure qpopper --with-pam and setup /etc/pam.d/pop3 to
>> use pam_mysql.so. This makes upgrades a bit less painful.
>>
>> Ken
>>
>>
>> Martin Kellermann wrote:
>>> hi list,
>>> is there a patch for qpopper 4.0.8 to support mysql auth?
>>> what i found is:
>>> http://www.asteroid-b612.org/software/qpopper-mysql/qpopper-mysql-0.14.patch
>>>
>>> but unfortunately it is for version 4.0.5 ...
>>> or, does it work for 4.0.8 ?
>>> i use qpopper for lots of years now....simply wonderful.
>>> but some mysql support would be fine...
>>> thanks for you hints..
>>> MK
>
>
Date: Mon, 15 Aug 2005 08:59:56 -0700
From: Ken A <ka at pacific dot net>
Subject: Re: qpopper 4.0.8 + mysql
A bit more info on a nss-mysql / pam-mysql setup.
http://www.linuxbourg.ch/cgi-bin/viewcvs.cgi/pam_mysqlim/Mini-HowToz
Ken A
Martin Kellermann wrote:
> hi Ken,
>
> i was too curious and tried qpopper/PAM/mysql.
> it is up and running... thanks!
>
> but there is one strange behaviour which i cant explain...
> users are ONLY authentivated if they have a local account (etc/passwd)!?
>
> so, maybe my question is a bit offtopic, but i think it is maybe a
> qpopper or config problem.
>
> case 1:
> - local user named "test" has local password "test"
> - mysql database has username "test" and password "otherthantest"
> - qpopper/pam authenticates the user correct when logging in with user
> "test" and pass "otherthantest"
> - pam_mysql syslogs without errors, everything is fine
>
> case 2:
> - no local user test exists
> - mysql database has username "test" and password "otherthantest"
> - qpopper/pam DOES NOT authenticate the user logging in with user "test"
> and pass "otherthantest"
> - pam_mysql syslogs shows NOTHING !
> it appears that, when there is no corresponding local user, the whole
> PAM mechanism isnt started...
>
> i configured qpopper with:
> ./configure --prefix=/usr --sysconfdir=/etc --enable-shy --with-pam=pop3
>
> my /etc/pam.conf is this:
> pop3 auth required pam_mysql.so verbose=1 host=localhost db=* table=*
> user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
> pop3 account required pam_mysql.so verbose=1 host=localhost db=* table=*
> user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
>
> any ideas?
> is there a way to tell qpopper to ONLY use PAM for authentication?
>
> thanks in advance
>
> MK
>
> At 00:36 13.08.2005, you wrote:
>> An alternative to patching qpopper is pam_mysql, which works fine with
>> 4.08. Just configure qpopper --with-pam and setup /etc/pam.d/pop3 to
>> use pam_mysql.so. This makes upgrades a bit less painful.
>>
>> Ken
>>
>>
>> Martin Kellermann wrote:
>>> hi list,
>>> is there a patch for qpopper 4.0.8 to support mysql auth?
>>> what i found is:
>>> http://www.asteroid-b612.org/software/qpopper-mysql/qpopper-mysql-0.14.patch
>>>
>>> but unfortunately it is for version 4.0.5 ...
>>> or, does it work for 4.0.8 ?
>>> i use qpopper for lots of years now....simply wonderful.
>>> but some mysql support would be fine...
>>> thanks for you hints..
>>> MK
>
>
Date: Wed, 17 Aug 2005 17:25:50 +0200
From: Martin Kellermann <Kellermann at sk-datentechnik dot com>
Subject: Re: qpopper 4.0.8 + mysql
thank you - and shame on me for not thinking about it...
of course you are right - changing uids is difficult for qpopper,
when no local uids exists ;-)
will try the NSS stuff later...time is running and it works now as expexted...
but there is another weird thing i didnt dare to ask:
when using the -t logfile option (debugging disabled) qpopper logs to the given
file instead of syslog. but i see blank lines between the entries...?
example:
Aug 17 17:14:39.768 2005 [20525] (v4.0.8) Servicing request from blabla
Aug 17 17:14:39.768 2005
Aug 17 17:14:39.829 2005 [20525] Stats: username 0 0 0 0 blabla
Aug 17 17:14:39.829 2005
Aug 17 17:14:40.968 2005 [20528] (v4.0.8) Servicing request from blabla
Aug 17 17:14:40.968 2005
Aug 17 17:14:41.013 2005 [20528] Stats: username 0 0 0 0 blabla
Aug 17 17:14:41.013 2005
huh? is this normal behaviour?
thanks again in advance
MK
At 17:42 15.08.2005, you wrote:
>Qpopper changes uids to read/write spools and temp drop files.
>Authentication is just one part of PAM. AFAIK, pam_mysql doesn't yet
>implement PAM 'account' functions, so you must have a local uid in
>/etc/passwd. The local user need not have a password. I simply
>create the local users with no home dir, no shell, no password, etc.
>
>Another possibility is using NSS to put uids in mysql, and PAM
>together, but I have not tried this. Here's an interesting doc that
>explains some of these issues.
>http://users.linuxbourg.ch/fvgoto/informatica/tbsc/doc/final/pam_mysqlim.pdf
>
>Ken A.
>
>
>Martin Kellermann wrote:
>>hi Ken,
>>i was too curious and tried qpopper/PAM/mysql.
>>it is up and running... thanks!
>>but there is one strange behaviour which i cant explain...
>>users are ONLY authentivated if they have a local account (etc/passwd)!?
>>so, maybe my question is a bit offtopic, but i think it is maybe a
>>qpopper or config problem.
>>case 1:
>>- local user named "test" has local password "test"
>>- mysql database has username "test" and password "otherthantest"
>>- qpopper/pam authenticates the user correct when logging in with
>>user "test" and pass "otherthantest"
>>- pam_mysql syslogs without errors, everything is fine
>>case 2:
>>- no local user test exists
>>- mysql database has username "test" and password "otherthantest"
>>- qpopper/pam DOES NOT authenticate the user logging in with user
>>"test" and pass "otherthantest"
>>- pam_mysql syslogs shows NOTHING !
>>it appears that, when there is no corresponding local user, the
>>whole PAM mechanism isnt started...
>>i configured qpopper with:
>>./configure --prefix=/usr --sysconfdir=/etc --enable-shy --with-pam=pop3
>>my /etc/pam.conf is this:
>>pop3 auth required pam_mysql.so verbose=1 host=localhost db=*
>>table=* user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
>>pop3 account required pam_mysql.so verbose=1 host=localhost db=*
>>table=* user=* passwd=* usercolumn=* passwdcolumn=* where=active=1 crypt=0
>>any ideas?
>>is there a way to tell qpopper to ONLY use PAM for authentication?
>>thanks in advance
>>MK
>>At 00:36 13.08.2005, you wrote:
>>>An alternative to patching qpopper is pam_mysql, which works fine
>>>with 4.08. Just configure qpopper --with-pam and setup
>>>/etc/pam.d/pop3 to use pam_mysql.so. This makes upgrades a bit less painful.
>>>
>>>Ken
>>>
>>>
>>>Martin Kellermann wrote:
>>>>hi list,
>>>>is there a patch for qpopper 4.0.8 to support mysql auth?
>>>>what i found is:
>>>>http://www.asteroid-b612.org/software/qpopper-mysql/qpopper-mysql-0.14.patch
>>>>but unfortunately it is for version 4.0.5 ...
>>>>or, does it work for 4.0.8 ?
>>>>i use qpopper for lots of years now....simply wonderful.
>>>>but some mysql support would be fine...
>>>>thanks for you hints..
>>>>MK
Date: Thu, 08 Sep 2005 13:36:49 +0200
From: "M dot Kellermann" <Kellermann at sk-datentechnik dot com>
Subject: empty lines in logfile
hi list,
since there was no response for over a month i'll try it again:
when using the -t logfile option (debugging disabled) qpopper logs to the given
file instead of syslog. but i see blank lines between the entries...?
example:
Aug 17 17:14:39.768 2005 [20525] (v4.0.8) Servicing request from blabla
Aug 17 17:14:39.768 2005
Aug 17 17:14:39.829 2005 [20525] Stats: username 0 0 0 0 blabla
Aug 17 17:14:39.829 2005
Aug 17 17:14:40.968 2005 [20528] (v4.0.8) Servicing request from blabla
Aug 17 17:14:40.968 2005
Aug 17 17:14:41.013 2005 [20528] Stats: username 0 0 0 0 blabla
Aug 17 17:14:41.013 2005
huh? is this normal behaviour?
qpopper is started via xinetd:
server = /usr/sbin/popper
server_args = qpopper -s -R -t /var/log/pop.log
logging works fine .... but where do these empty lines with just
the timestamp and no logging text come from?
is noone but me logging to a separate file?
i cant believe that this is a qpopper bug...
thanks again in advance
MK
Date: Thu, 08 Sep 2005 10:21:56 -0400
From: Daniel Senie <dts at senie dot com>
Subject: Re: empty lines in logfile
At 07:36 AM 9/8/2005, M. Kellermann wrote:
>hi list,
>
>since there was no response for over a month i'll try it again:
>
>when using the -t logfile option (debugging disabled) qpopper logs
>to the given
>file instead of syslog. but i see blank lines between the entries...?
>example:
>Aug 17 17:14:39.768 2005 [20525] (v4.0.8) Servicing request from blabla
>Aug 17 17:14:39.768 2005
>Aug 17 17:14:39.829 2005 [20525] Stats: username 0 0 0 0 blabla
>Aug 17 17:14:39.829 2005
>Aug 17 17:14:40.968 2005 [20528] (v4.0.8) Servicing request from blabla
>Aug 17 17:14:40.968 2005
>Aug 17 17:14:41.013 2005 [20528] Stats: username 0 0 0 0 blabla
>Aug 17 17:14:41.013 2005
>
>huh? is this normal behaviour?
The code in the function vlogit() in common/logit.c explicitly does
this. I have no idea WHY it explicitly does this, but it does. You'll
find there two fprintf() calls followed by an fflush(). The second
fprintf is the issue.
Anyone know why? Should we just remove the second fprintf?
>qpopper is started via xinetd:
> server = /usr/sbin/popper
> server_args = qpopper -s -R -t /var/log/pop.log
>
>logging works fine .... but where do these empty lines with just
>the timestamp and no logging text come from?
>
>is noone but me logging to a separate file?
>i cant believe that this is a qpopper bug...
>
>thanks again in advance
>
>MK
Date: Thu, 08 Sep 2005 18:07:23 +0200
From: "M dot Kellermann" <kellermann at sk-datentechnik dot com>
Subject: Re: empty lines in logfile
At 16:21 08.09.2005, you wrote:
>At 07:36 AM 9/8/2005, M. Kellermann wrote:
>>hi list,
>>
>>since there was no response for over a month i'll try it again:
>>
>>when using the -t logfile option (debugging disabled) qpopper logs
>>to the given
>>file instead of syslog. but i see blank lines between the entries...?
>>example:
>>Aug 17 17:14:39.768 2005 [20525] (v4.0.8) Servicing request from blabla
>>Aug 17 17:14:39.768 2005
>>Aug 17 17:14:39.829 2005 [20525] Stats: username 0 0 0 0 blabla
>>Aug 17 17:14:39.829 2005
>>Aug 17 17:14:40.968 2005 [20528] (v4.0.8) Servicing request from blabla
>>Aug 17 17:14:40.968 2005
>>Aug 17 17:14:41.013 2005 [20528] Stats: username 0 0 0 0 blabla
>>Aug 17 17:14:41.013 2005
>>
>>huh? is this normal behaviour?
>
>The code in the function vlogit() in common/logit.c explicitly does
>this. I have no idea WHY it explicitly does this, but it does.
>You'll find there two fprintf() calls followed by an fflush(). The
>second fprintf is the issue.
>
>Anyone know why? Should we just remove the second fprintf?
i tried removing the second fprintf, and now my logit.c looks like this:
.....
date_time = ctime ( (time_t *) &tval.tv_sec );
pDate = date_time + 4; /* skip day of week name */
iDateLen = strlen ( pDate ); /* length including year */
pYear = pDate + ( iDateLen - 5 ); /* point to start of year */
date_time [ iDateLen - 2 ] = '\0'; /* cut off the year */
lMsec = (tval.tv_usec + 500) / 1000; /* convert useconds
to milliseconds */
fprintf ( str, "%s.%03ld %.4s [%ld] %s\n",
pDate, lMsec, pYear, (long) getpid(), msgbuf );
fflush ( str );
}
else {
syslog ( loglev, "%s", msgbuf) ;
}
if ( iChunk == -1 ) {
/*
* We blew out the format buffer.
*/
if ( str ) {
fprintf ( str, "%s [%ld] Buffer size exceeded logging msg: %s\n",
date_time, (long) getpid(), format );
fflush ( str );
}
else {
.....
but it doesnt solve the blank lines.... ;(
C has never been my friend....but, doesnt the \n in the fprintf statement
mean "new line" ??
any programmers hints welcome...
thanks
MK
>>qpopper is started via xinetd:
>> server = /usr/sbin/popper
>> server_args = qpopper -s -R -t /var/log/pop.log
>>
>>logging works fine .... but where do these empty lines with just
>>the timestamp and no logging text come from?
>>
>>is noone but me logging to a separate file?
>>i cant believe that this is a qpopper bug...
>>
>>thanks again in advance
>>
>>MK
Date: Thu, 08 Sep 2005 12:58:20 -0400
From: Daniel Senie <dts at senie dot com>
Subject: Re: empty lines in logfile
At 12:07 PM 9/8/2005, M. Kellermann wrote:
>At 16:21 08.09.2005, you wrote:
>>At 07:36 AM 9/8/2005, M. Kellermann wrote:
>>>hi list,
>>>
>>>since there was no response for over a month i'll try it again:
>>>
>>>when using the -t logfile option (debugging disabled) qpopper logs
>>>to the given
>>>file instead of syslog. but i see blank lines between the entries...?
>>>example:
>>>Aug 17 17:14:39.768 2005 [20525] (v4.0.8) Servicing request from blabla
>>>Aug 17 17:14:39.768 2005
>>>Aug 17 17:14:39.829 2005 [20525] Stats: username 0 0 0 0 blabla
>>>Aug 17 17:14:39.829 2005
>>>Aug 17 17:14:40.968 2005 [20528] (v4.0.8) Servicing request from blabla
>>>Aug 17 17:14:40.968 2005
>>>Aug 17 17:14:41.013 2005 [20528] Stats: username 0 0 0 0 blabla
>>>Aug 17 17:14:41.013 2005
>>>
>>>huh? is this normal behaviour?
>>
>>The code in the function vlogit() in common/logit.c explicitly does
>>this. I have no idea WHY it explicitly does this, but it does.
>>You'll find there two fprintf() calls followed by an fflush(). The
>>second fprintf is the issue.
>>
>>Anyone know why? Should we just remove the second fprintf?
>
>i tried removing the second fprintf, and now my logit.c looks like this:
>.....
> date_time = ctime ( (time_t *) &tval.tv_sec );
> pDate = date_time + 4; /* skip day of week name */
> iDateLen = strlen ( pDate ); /* length including year */
> pYear = pDate + ( iDateLen - 5 ); /* point to start of year */
> date_time [ iDateLen - 2 ] = '\0'; /* cut off the year */
> lMsec = (tval.tv_usec + 500) / 1000; /* convert
> useconds to milliseconds */
> fprintf ( str, "%s.%03ld %.4s [%ld] %s\n",
> pDate, lMsec, pYear, (long) getpid(), msgbuf );
> fflush ( str );
> }
> else {
> syslog ( loglev, "%s", msgbuf) ;
> }
>
> if ( iChunk == -1 ) {
> /*
> * We blew out the format buffer.
> */
> if ( str ) {
> fprintf ( str, "%s [%ld] Buffer size exceeded logging msg: %s\n",
> date_time, (long) getpid(), format );
> fflush ( str );
> }
> else {
>.....
>
>but it doesnt solve the blank lines.... ;(
You're sure it recompiled and you placed the new executable in place
of the old one?
>C has never been my friend....but, doesnt the \n in the fprintf statement
>mean "new line" ??
It does, however if you were to remove the \n from the remaining
fprintf, then all your output would wind up on one line, with no line
ends. The \n on Unix systems puts out a linefeed, which is a line
end. On DOS/Windows systems it'd put out a CR/LF. You really do need that \n.
The fact is you're getting more than just blank lines, you're getting
lines with timestamp and other info. That info is coming from the
second fprintf, thus my question about whether the executable got
built right and is really in use now.
>any programmers hints welcome...
>
>
>thanks
>
>MK
>
>
>
>>>qpopper is started via xinetd:
>>> server = /usr/sbin/popper
>>> server_args = qpopper -s -R -t /var/log/pop.log
>>>
>>>logging works fine .... but where do these empty lines with just
>>>the timestamp and no logging text come from?
>>>
>>>is noone but me logging to a separate file?
>>>i cant believe that this is a qpopper bug...
>>>
>>>thanks again in advance
>>>
>>>MK
Date: Thu, 8 Sep 2005 16:06:24 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: empty lines in logfile
At 10:21 AM -0400 9/8/05, Daniel Senie wrote:
> The code in the function vlogit() in common/logit.c explicitly does
> this. I have no idea WHY it explicitly does this, but it does.
> You'll find there two fprintf() calls followed by an fflush(). The
> second fprintf is the issue.
>
> Anyone know why? Should we just remove the second fprintf?
The code has been deliberately written that way for many, many years.
I always assumed it was there for better readability, but that was
just an assumption.
Personally, I don't mind it either way. I make extensive use of
logging to a separate file, and am used to the extra lines. Perhaps
a configuration file option?
At 12:58 PM -0400 9/8/05, Daniel Senie wrote:
> You're sure it recompiled and you placed the new executable in
> place of the old one?
I make it a habit to tweak the version or banner information when I
make a change, in order to be able to verify that I'm running the new
code. It's too easy to run the old code.
--
Randall Gellens
Opinions are personal; facts are suspect; I speak for myself only
-------------- Randomly-selected tag: ---------------
I worry that the person who thought up Muzak may be thinking up
something else.
--Lily Tomlin
Date: Fri, 09 Sep 2005 08:50:19 +0200
From: "M dot Kellermann" <kellermann at sk-datentechnik dot com>
Subject: Re: empty lines in logfile
At 18:58 08.09.2005, you wrote:
>>...
>>i tried removing the second fprintf, and now my logit.c looks like this:
>>.....
>> date_time = ctime ( (time_t *) &tval.tv_sec );
>> pDate = date_time + 4; /* skip day of week name */
>> iDateLen = strlen ( pDate ); /* length including year */
>> pYear = pDate + ( iDateLen - 5 ); /* point to start of year */
>> date_time [ iDateLen - 2 ] = '\0'; /* cut off the year */
>> lMsec = (tval.tv_usec + 500) / 1000; /* convert
>> useconds to milliseconds */
>> fprintf ( str, "%s.%03ld %.4s [%ld] %s\n",
>> pDate, lMsec, pYear, (long) getpid(), msgbuf );
>> fflush ( str );
>> }
>> else {
>> syslog ( loglev, "%s", msgbuf) ;
>> }
>>
>> if ( iChunk == -1 ) {
>> /*
>> * We blew out the format buffer.
>> */
>> if ( str ) {
>> fprintf ( str, "%s [%ld] Buffer size exceeded logging
>> msg: %s\n",
>> date_time, (long) getpid(), format );
>> fflush ( str );
>> }
>> else {
>>.....
>>
>>but it doesnt solve the blank lines.... ;(
>
>You're sure it recompiled and you placed the new executable in place
>of the old one?
yep!
i changed the version number to 4.0.9!!
believe me - the blank lines are still there?!?
i have no idea why this is happening....
maybe the logit.c code isnt used at all!?
>The fact is you're getting more than just blank lines, you're
>getting lines with timestamp and other info. That info is coming
>from the second fprintf, thus my question about whether the
>executable got built right and is really in use now.
im absolutely sure, that the new compiled version is in use:
Sep 9 08:47:29.136 2005 [27225] (v4.0.9) Servicing request from "---" at ---
Sep 9 08:47:29.136 2005
Sep 9 08:47:29.552 2005 [27225] Stats: xxx 0 0 0 0 --- ---
Sep 9 08:47:29.552 2005
Sep 9 08:48:12.535 2005 [27229] (v4.0.9) Servicing request from "---" at ---
Sep 9 08:48:12.535 2005
Sep 9 08:48:12.622 2005 [27229] Stats: xxx 0 0 0 0 --- ---
Sep 9 08:48:12.622 2005
VERY strange....
Date: Fri, 09 Sep 2005 09:04:20 +0200
From: "M dot Kellermann" <kellermann at sk-datentechnik dot com>
Subject: [solved] Re: empty lines in logfile
ok, i found it.
the code in commen/logit.c is never used.
instead logging is done by popper/pop_log.c - and
removing the obsolete fprintf's works now fine for me.
maybe a configuration option to set this at runtime would be useful!?
thanks for all the help
MK
>Date: Fri, 09 Sep 2005 08:50:19 +0200
>To: qpopper@lists.pensive.org
>From: "M. Kellermann" <kellermann@sk-datentechnik.com>
>Subject: Re: empty lines in logfile
>
>At 18:58 08.09.2005, you wrote:
>>>...
>>>i tried removing the second fprintf, and now my logit.c looks like this:
>>>.....
>>> date_time = ctime ( (time_t *) &tval.tv_sec );
>>> pDate = date_time + 4; /* skip day of week name */
>>> iDateLen = strlen ( pDate ); /* length including year */
>>> pYear = pDate + ( iDateLen - 5 ); /* point to start of year */
>>> date_time [ iDateLen - 2 ] = '\0'; /* cut off the year */
>>> lMsec = (tval.tv_usec + 500) / 1000; /* convert
>>> useconds to milliseconds */
>>> fprintf ( str, "%s.%03ld %.4s [%ld] %s\n",
>>> pDate, lMsec, pYear, (long) getpid(), msgbuf );
>>> fflush ( str );
>>> }
>>> else {
>>> syslog ( loglev, "%s", msgbuf) ;
>>> }
>>>
>>> if ( iChunk == -1 ) {
>>> /*
>>> * We blew out the format buffer.
>>> */
>>> if ( str ) {
>>> fprintf ( str, "%s [%ld] Buffer size exceeded logging
>>> msg: %s\n",
>>> date_time, (long) getpid(), format );
>>> fflush ( str );
>>> }
>>> else {
>>>.....
>>>
>>>but it doesnt solve the blank lines.... ;(
>>
>>You're sure it recompiled and you placed the new executable in
>>place of the old one?
>
>yep!
>i changed the version number to 4.0.9!!
>believe me - the blank lines are still there?!?
>
>i have no idea why this is happening....
>maybe the logit.c code isnt used at all!?
>
>
>>The fact is you're getting more than just blank lines, you're
>>getting lines with timestamp and other info. That info is coming
>>from the second fprintf, thus my question about whether the
>>executable got built right and is really in use now.
>
>im absolutely sure, that the new compiled version is in use:
>
>Sep 9 08:47:29.136 2005 [27225] (v4.0.9) Servicing request from "---" at ---
>Sep 9 08:47:29.136 2005
>Sep 9 08:47:29.552 2005 [27225] Stats: xxx 0 0 0 0 --- ---
>Sep 9 08:47:29.552 2005
>Sep 9 08:48:12.535 2005 [27229] (v4.0.9) Servicing request from "---" at ---
>Sep 9 08:48:12.535 2005
>Sep 9 08:48:12.622 2005 [27229] Stats: xxx 0 0 0 0 --- ---
>Sep 9 08:48:12.622 2005
>
>VERY strange....
Date: Mon, 12 Sep 2005 08:18:16 -0700
From: Ken A <ka at pacific dot net>
Subject: qpopper ***glibc detected *** malloc() : memory corruption
Some end users are reporting occassional error messages in Outlook Express:
There was a problem logging onto your mail server. Your Password was
rejected. Account: 'BLAH', Server: 'server.host.tld', Protocol: POP3,
Server Response: '*** glibc detected *** malloc(): memory corruption:
0x08081740 ***', Port: 110, Secure(SSL): No, Server Error: 0x800CCC90,
Error Number: 0x800CCC92
Googling on this:
"***glibc detected *** malloc() : memory corruption"
gives me various reports of bugs in software - memory not being properly
freed, etc.. This reporting is new in recent versions of glibc.
Any ideas on where in qpopper I might be able to turn off this error
message, or at least direct it to the logfiles, rather than the client!
Or how I should report this bug, if it is a bug in qpopper?
Thanks,
Ken A
Date: Mon, 12 Sep 2005 08:29:52 -0700
From: Ken A <ka at pacific dot net>
Subject: re: qpopper ***glibc detected *** malloc() : memory corruption
Additional info:
The error " ***glibc detected *** malloc() : memory corruption" is
coming up in a pop3 client running windows. Server is running qpopper
4.0.8, on fedora core 3 with latest patches.
Below is some info about the error message and a hint about how to
suppress this error message.
So, it looks like putting "MALLOC_CHECK_ 0" into the qpopper init file
would suppress the error message, but there seems to be a bug in the
qpopper code?
Thanks,
Ken
From 'man malloc'
> Recent versions of Linux libc (later than 5.4.23) and GNU libc (2.x)
> include a malloc implementation which is tunable via environment
> variables. When MALLOC_CHECK_ is set, a special (less efficient)
> implementation is used which is designed to be tolerant against
> simple errors, such as double calls of free() with the same
> argument, or overruns of a single byte (off-by-one bugs). Not all
> such errors can be protected against, however, and memory leaks can
> result. If MALLOC_CHECK_ is set to 0, any detected heap
> corruption is silently ignored; if set to 1, a diagnostic is printed
> on stderr; if set to 2, abort() is called immediately. This can be
> useful because otherwise a crash may happen much later, and the true
> cause for the problem is then very hard to track down.
Date: Wed, 14 Sep 2005 10:17:17 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: re: qpopper ***glibc detected *** malloc() : memory corruption
At 8:29 AM -0700 9/12/05, Ken A wrote:
> Additional info:
>
> The error " ***glibc detected *** malloc() : memory corruption" is
> coming up in a pop3 client running windows. Server is running
> qpopper 4.0.8, on fedora core 3 with latest patches.
>
> Below is some info about the error message and a hint about how to
> suppress this error message.
>
> So, it looks like putting "MALLOC_CHECK_ 0" into the qpopper init
> file would suppress the error message, but there seems to be a bug
> in the qpopper code?
Can you reproduce this? If so, would it be possible to do so running
Qpopper under truss(1) or ktrace(1) or whatever the equivalent is on
your platform?
If there is a bug it would be good to track it down and fix it.
>
> Thanks,
> Ken
>
>
> From 'man malloc'
>
>> Recent versions of Linux libc (later than 5.4.23) and GNU libc (2.x)
>> include a malloc implementation which is tunable via environment
>> variables. When MALLOC_CHECK_ is set, a special (less efficient)
>> implementation is used which is designed to be tolerant against
>> simple errors, such as double calls of free() with the same
>> argument, or overruns of a single byte (off-by-one bugs). Not all
>> such errors can be protected against, however, and memory leaks can
>> result. If MALLOC_CHECK_ is set to 0, any detected heap
>> corruption is silently ignored; if set to 1, a diagnostic is printed
>> on stderr; if set to 2, abort() is called immediately. This can be
>> useful because otherwise a crash may happen much later, and the true
>> cause for the problem is then very hard to track down.
--
Randall Gellens
Opinions are personal; facts are suspect; I speak for myself only
-------------- Randomly-selected tag: ---------------
The significant problems that we face today cannot be solved at the
same level of thinking that created them. --Albert Einstein
Date: Wed, 14 Sep 2005 11:01:14 -0700
From: Ken A <ka at pacific dot net>
Subject: Re: qpopper ***glibc detected *** malloc() : memory corruption
Randall Gellens wrote:
> At 8:29 AM -0700 9/12/05, Ken A wrote:
>
>> Additional info:
>>
>> The error " ***glibc detected *** malloc() : memory corruption" is
>> coming up in a pop3 client running windows. Server is running qpopper
>> 4.0.8, on fedora core 3 with latest patches.
>>
>> Below is some info about the error message and a hint about how to
>> suppress this error message.
>>
>> So, it looks like putting "MALLOC_CHECK_ 0" into the qpopper init
>> file would suppress the error message, but there seems to be a bug in
>> the qpopper code?
>
> Can you reproduce this? If so, would it be possible to do so running
> Qpopper under truss(1) or ktrace(1) or whatever the equivalent is on
> your platform?
>
> If there is a bug it would be good to track it down and fix it.
It's a busy production machine, and only a couple users have reported
it. The condition that seems to be common with the reports is checking
pop3 mail 2 times quickly in succession. I have not been able to
reproduce it. One user had Outlook Express set to check the same pop3
box twice, and the other was on a high latency satellite link and was
impatiently clicking send & receive.
If I get some time, I'll setup a test popper on another box and see if I
can repeat it.
Thanks,
Ken A
Pacific.Net
>
>>
>> Thanks,
>> Ken
>>
>>
>> From 'man malloc'
>>
>>> Recent versions of Linux libc (later than 5.4.23) and GNU libc (2.x)
>>> include a malloc implementation which is tunable via environment
>>> variables. When MALLOC_CHECK_ is set, a special (less efficient)
>>> implementation is used which is designed to be tolerant against
>>> simple errors, such as double calls of free() with the same
>>> argument, or overruns of a single byte (off-by-one bugs). Not all
>>> such errors can be protected against, however, and memory leaks can
>>> result. If MALLOC_CHECK_ is set to 0, any detected heap
>>> corruption is silently ignored; if set to 1, a diagnostic is printed
>>> on stderr; if set to 2, abort() is called immediately. This can be
>>> useful because otherwise a crash may happen much later, and the true
>>> cause for the problem is then very hard to track down.
>
>
From: "GioiaBa" <gioiaba at slacky dot it>
Subject: error flushing output to client
Date: Mon, 19 Sep 2005 14:56:30 +0200
Hi list !
my problem is that I have these entries in my /var/log/maillog file
#############################################################
Sep 18 10:31:36 mail popper[13054]: I/O error flushing output to client
user1 at adsl000.domain.com.au [xxx.xxx.xxx.xxx]: Operation not permitted
(1)
#############################################################
and the client itself (using Oexpress or Outlook) gets an error that it
disappears when it do 'send and receive' once again..
it happens not all days, and not for all the users..
I searched the list, and found someonelse having the same trouble, but I
can't still have a solution...
I'm running slackware 10 with
log2db-0.6 ; qpopper4.0.5 ; sendmail 8.12.11
thanks anyone..
Date: Mon, 19 Sep 2005 09:33:00 -0400
From: Daniel Senie <dts at senie dot com>
Subject: Re: error flushing output to client
At 08:56 AM 9/19/2005, GioiaBa wrote:
>Hi list !
>my problem is that I have these entries in my /var/log/maillog file
>#############################################################
>Sep 18 10:31:36 mail popper[13054]: I/O error flushing output to
>client user1 at adsl000.domain.com.au [xxx.xxx.xxx.xxx]: Operation
>not permitted (1)
This error indicates the TCP session has gone away from under
qpopper. In other words, the TCP session was closed because the
remote endpoint (PC in this case) closed the connection.
>#############################################################
>and the client itself (using Oexpress or Outlook) gets an error that
>it disappears when it do 'send and receive' once again..
>it happens not all days, and not for all the users..
>I searched the list, and found someonelse having the same trouble,
>but I can't still have a solution...
>I'm running slackware 10 with
>log2db-0.6 ; qpopper4.0.5 ; sendmail 8.12.11
>thanks anyone..
Generally this is caused by timeouts that are too short on the PC, or
antivirus software doing dumb things, with the latter becoming
epidemic of late. We've moved a lot of customers to POP on alternate
port with TLS (TCP/995) to get their email away from the interference
caused by the latest live updates to Norton Antivirus 2005. There may
be a better solution to dealing with Norton (configuration, bug fix,
or uninstall) but this work-around has at least made it possible to
keep customers functional. Your problem may or may not be at all
related to antivirus, but it's something to look at.
From: "GioiaBa" <gioiaba at slacky dot it>
Subject: Re: error flushing output to client
Date: Mon, 19 Sep 2005 16:41:45 +0200
Thanks Daniel
well in fact, I don't konw the pc user environment for all users having
these trouble sometimes..
but I know for sure that Some of those, do NOT have Norton Antivirus 2005
installed.. so there has to be another reason for this..
but I know it is not simple to find out why it happens..
any other hints will be appreciated..
thanks again..
: )
Daniel Senie Scrive:
> At 08:56 AM 9/19/2005, GioiaBa wrote:
>> Hi list !
>> my problem is that I have these entries in my /var/log/maillog file
>> #############################################################
>> Sep 18 10:31:36 mail popper[13054]: I/O error flushing output to client
>> user1 at adsl000.domain.com.au [xxx.xxx.xxx.xxx]: Operation not permitted
>> (1)
>
> This error indicates the TCP session has gone away from under qpopper. In
> other words, the TCP session was closed because the remote endpoint (PC in
> this case) closed the connection.
>
>> #############################################################
>> and the client itself (using Oexpress or Outlook) gets an error that it
>> disappears when it do 'send and receive' once again..
>> it happens not all days, and not for all the users..
>> I searched the list, and found someonelse having the same trouble, but I
>> can't still have a solution...
>> I'm running slackware 10 with
>> log2db-0.6 ; qpopper4.0.5 ; sendmail 8.12.11
>> thanks anyone..
>
> Generally this is caused by timeouts that are too short on the PC, or
> antivirus software doing dumb things, with the latter becoming epidemic of
> late. We've moved a lot of customers to POP on alternate port with TLS
> (TCP/995) to get their email away from the interference caused by the
> latest live updates to Norton Antivirus 2005. There may be a better
> solution to dealing with Norton (configuration, bug fix, or uninstall) but
> this work-around has at least made it possible to keep customers
> functional. Your problem may or may not be at all related to antivirus,
> but it's something to look at.
>
GB
Date: Mon, 19 Sep 2005 12:19:26 -0400 (EDT)
From: Alan Brown <alanb at digistar dot com>
Subject: Re: error flushing output to client
On Mon, 19 Sep 2005, GioiaBa wrote:
> but I know for sure that Some of those, do NOT have Norton Antivirus 2005
> installed.. so there has to be another reason for this..
> but I know it is not simple to find out why it happens..
large messages, slow connections, short timeouts set on the clients
(default is usually 60 seconds per message)
From: "Doug Wellington" <ddw at rinconresearch dot com>
Subject: "Password supplied is incorrect" error with Solaris 9...
Date: Fri, 23 Sep 2005 16:31:33 -0700
I'm not able to login to Qpopper 4.0.8 running on a Solaris 9 server. I
keep getting the error:
-ERR [AUTH] Password supplied for "ddw" is incorrect
I can telnet and ftp to the server, but qpopper won't let me in. I looked
in the FAQ and did a google search, but the only thing I've found so far is
to use --enable-specialauth. I have done a make realclean and configured
both with and without --enable-specialauth and the binary ends up being
exactly the same. Is there an issue with which version of gcc I compile
with? (I'm using 3.4.2...)
Thanks,
Doug
Date: Fri, 23 Sep 2005 15:20:08 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: ***spam*** "Password supplied is incorrect" error with Solaris 9...
On Fri, Sep 23, 2005 at 04:31:33PM -0700, Doug Wellington wrote:
> I'm not able to login to Qpopper 4.0.8 running on a Solaris 9 server. I
> keep getting the error:
>
> -ERR [AUTH] Password supplied for "ddw" is incorrect
>
> I can telnet and ftp to the server, but qpopper won't let me in. I looked
> in the FAQ and did a google search, but the only thing I've found so far is
> to use --enable-specialauth. I have done a make realclean and configured
> both with and without --enable-specialauth and the binary ends up being
> exactly the same. Is there an issue with which version of gcc I compile
> with? (I'm using 3.4.2...)
This is perhaps a dumb question, but you are running qpopper as the
root user, right? It needs that to be able to check the password and
then change uid.
If so, then on Solaris I would check for a problem with PAM
configuration.
-- Clifton
--
Clifton Royston -- cliftonr@tikitechnologies.com
Tiki Technologies Lead Programmer/Software Architect
"My own personal theory is that this is the very dawn of the world.
We're hardly more than an eyeblink away from the fall of Troy, and
scarcely an interglaciation removed from the Altamira cave painters. We
live in extremely interesting ancient times.
I like this idea. It encourages us to be earnest and ingenious and
brave, as befits ancestral peoples; but keeps us from deciding that
because we don't know all the answers, they must be unknowable and thus
unprofitable to pursue." -- Teresa Nielsen Hayden, 1995
Date: Mon, 26 Sep 2005 12:48:45 -0500
From: Jerry K <qpopper at oryx dot cc>
Subject: Re: "Password supplied is incorrect" error with Solaris 9...
This is my configure string that I use for my solaris 9 systems. I came up with
this based on other hints that I have read on this list, plus reading the docs.
The enable-temp-drop-dir works well for me as I use user quotas.
I am using the same version of gcc that you are.
Hope this helps,
Jerry Kemp
./configure --disable-apop \
--enable-specialauth \
--enable-uw-kludge \
--disable-check-pwmax \
--enable-log-login \
--without-pam \
--enable-temp-drop-dir=/var/tmp
Doug Wellington wrote:
> I'm not able to login to Qpopper 4.0.8 running on a Solaris 9 server. I
> keep getting the error:
>
> -ERR [AUTH] Password supplied for "ddw" is incorrect
>
> I can telnet and ftp to the server, but qpopper won't let me in. I looked
> in the FAQ and did a google search, but the only thing I've found so far is
> to use --enable-specialauth. I have done a make realclean and configured
> both with and without --enable-specialauth and the binary ends up being
> exactly the same. Is there an issue with which version of gcc I compile
> with? (I'm using 3.4.2...)
>
> Thanks,
> Doug
>
From: "Edward Chase" <echase at studentweb dot providence dot edu>
Subject: qpopper and quotas
Date: Mon, 26 Sep 2005 15:04:09 -0400
I'm having issues with my users blowing their storage quotas while checking
their mail.
Consider this...
User is at storage limit.
User checks mail.
/var/mail/user gets copied to /var/spool/poptemp/.user.pop
New mail comes into /var/mail/user
System cannot combine the 2 files back because it exceeds storage limit for
user.
Obviously user is leaving mail behind on server...
I haven't figured out a way around this.
Anyone run into this and figure out a way around it?
Thanks.
--
Edward Chase
Providence College
Information Technology
Date: Mon, 26 Sep 2005 15:31:23 -0400
From: Daniel Senie <dts at senie dot com>
Subject: Re: qpopper and quotas
At 03:04 PM 9/26/2005, Edward Chase wrote:
>I'm having issues with my users blowing their storage quotas while checking
>their mail.
>
>Consider this...
>
>User is at storage limit.
>User checks mail.
>/var/mail/user gets copied to /var/spool/poptemp/.user.pop
>New mail comes into /var/mail/user
>System cannot combine the 2 files back because it exceeds storage limit for
>user.
>
>Obviously user is leaving mail behind on server...
>
>I haven't figured out a way around this.
>
>Anyone run into this and figure out a way around it?
The only solutions I've seen involve putting the temp storage on a
separate machine.
The real answer, IMO, and one I haven't chased yet is to have the
SMTP server (sendmail, etc.) handle mailbox sizing issues, and bounce
messages when people are over quota. It should be possible to
integrate such without that much effort, but the sendmail/procmail
universe seems to lack such (I'm certainly open to pointers if folks
have solutions to this).
Ultimately, I really don't want to use the file system quota
mechanism to deal with limiting users' mailbox sizes. Perhaps quotas
work for you (other than the obvious qpopper question you asked
about) but they seem like the wrong tool for the job.
Date: Mon, 26 Sep 2005 12:28:46 -0700
From: Ken A <ka at pacific dot net>
Subject: Re: qpopper and quotas
Are /var/mail and /var/spool/poptemp on the same partition?
Are you running procmail as LDA ?
It respects quotas.
'mail' does not.
Ken
Edward Chase wrote:
> I'm having issues with my users blowing their storage quotas while checking
> their mail.
>
> Consider this...
>
> User is at storage limit.
> User checks mail.
> /var/mail/user gets copied to /var/spool/poptemp/.user.pop
> New mail comes into /var/mail/user
> System cannot combine the 2 files back because it exceeds storage limit for
> user.
>
> Obviously user is leaving mail behind on server...
>
> I haven't figured out a way around this.
>
> Anyone run into this and figure out a way around it?
>
> Thanks.
>
Date: Mon, 26 Sep 2005 10:02:54 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: qpopper and quotas
On Mon, Sep 26, 2005 at 03:04:09PM -0400, Edward Chase wrote:
> I'm having issues with my users blowing their storage quotas while checking
> their mail.
>
> Consider this...
>
> User is at storage limit.
> User checks mail.
> /var/mail/user gets copied to /var/spool/poptemp/.user.pop
> New mail comes into /var/mail/user
> System cannot combine the 2 files back because it exceeds storage limit for
> user.
>
> Obviously user is leaving mail behind on server...
>
> I haven't figured out a way around this.
>
> Anyone run into this and figure out a way around it?
Yes, common known issue which admins have been batting around for
years, with no entirely satisfactory solution.
Do you currently have /var/spool/poptemp on the same file system as
/var/mail? That exacerbates it.
Generally AFAIK the best you can do is:
* Run qpopper in server mode which optimizes away some spool copy
operations;
* Put /var/spool/poptemp on a separate FS;
* Keep hard quota set to at least 2x soft quota (so that they can
briefly spike to 2x the normal quota;
* Set up a separate script which runs nightly and informs the users
they are near to going over their (soft) quota, and to clean up
their act!
-- Clifton
--
Clifton Royston -- cliftonr@tikitechnologies.com
Tiki Technologies Lead Programmer/Software Architect
"My own personal theory is that this is the very dawn of the world.
We're hardly more than an eyeblink away from the fall of Troy, and
scarcely an interglaciation removed from the Altamira cave painters. We
live in extremely interesting ancient times.
I like this idea. It encourages us to be earnest and ingenious and
brave, as befits ancestral peoples; but keeps us from deciding that
because we don't know all the answers, they must be unknowable and thus
unprofitable to pursue." -- Teresa Nielsen Hayden, 1995
Date: Mon, 26 Sep 2005 14:44:28 -0700
From: Ken A <ka at pacific dot net>
Subject: Re: qpopper and quotas
Daniel Senie wrote:
> At 03:04 PM 9/26/2005, Edward Chase wrote:
>> I'm having issues with my users blowing their storage quotas while
>> checking
>> their mail.
>>
>> Consider this...
>>
>> User is at storage limit.
>> User checks mail.
>> /var/mail/user gets copied to /var/spool/poptemp/.user.pop
>> New mail comes into /var/mail/user
>> System cannot combine the 2 files back because it exceeds storage
>> limit for
>> user.
>>
>> Obviously user is leaving mail behind on server...
>>
>> I haven't figured out a way around this.
>>
>> Anyone run into this and figure out a way around it?
>
> The only solutions I've seen involve putting the temp storage on a
> separate machine.
>
> The real answer, IMO, and one I haven't chased yet is to have the SMTP
> server (sendmail, etc.) handle mailbox sizing issues, and bounce
> messages when people are over quota. It should be possible to integrate
> such without that much effort, but the sendmail/procmail universe seems
> to lack such (I'm certainly open to pointers if folks have solutions to
> this).
>
> Ultimately, I really don't want to use the file system quota mechanism
> to deal with limiting users' mailbox sizes. Perhaps quotas work for you
> (other than the obvious qpopper question you asked about) but they seem
> like the wrong tool for the job.
>
>
The real bugger is fastupdate w/quotas. It is nice feature, and speeds
up qpopper quite a bit, but depends on being on the same partition,
which means quotas won't work, since a user can't even check pop mail if
they are over quota, since qpopper can't create a temp file. procmail
will work, but only if temp drop is on a different partition that
doesn't enforce quotas.
I've found it best to stay away from hard quotas and just use a
combination of notification and/or mailbox pruning mechanism.
One helpful script is expire_mail.pl:
http://www.cpan.org/scripts/mailstuff/expire_mail.pl
Ken A
Date: Mon, 26 Sep 2005 15:35:36 -0500
From: David Champion <dgc at uchicago dot edu>
Subject: Re: qpopper and quotas
* On 2005.09.26, in <159254715904826091923@lists.pensive.org>,
* "Daniel Senie" <dts@senie.com> wrote:
>
> Ultimately, I really don't want to use the file system quota
> mechanism to deal with limiting users' mailbox sizes. Perhaps quotas
> work for you (other than the obvious qpopper question you asked
> about) but they seem like the wrong tool for the job.
That was our conclusion. We ran sendmail, and had a script which
nightly ran through all users' mailboxes (both the shared IMAP/POP/NFS
inbox and the IMAP folder hierarchy). We had a colon-delimited file
listing people's quotas, with a special entry for the default quota.
Another CDF listed people who were known to be over quota already, and
indicated how many days they had until quota was enforced.
Any time the script found someone to be over quota, it looked in the
second CDF. If the user was found, then their day cound would be
decremented. If not, they'd be added with the maximum number of days.
Once a user hit zero days, the script formulated an entry for sendmail's
access file blocking new delivery. I believe these were stored in a
separate file, and a Makefile in /etc/mail compiled all our various
access files into a single access map.
It wasn't hard to put together, really. Unfortunately we don't still
run this, so I can only describe it. It worked pretty well, but we gave
it up when we saw the light of the Mail God and purchased an enterprise
tur[n]key mail solution which solves all our problems in less time, with
less labor, and at lower expense -- except when it doesn't, which is
most of the time.
If I had it to do again I think I'd make it a constantly-running
process. Rather than running nightly, it would run again as soon as it
finished a cycle. Instead of storing quota days in the CDF, it would
mark timestamps that a user was first and most recently observed to
be over quota, and a separate script would periodically (more often
than daily) update /etc/mail/access from the difference between those
timestamps.
--
-D. dgc@uchicago.edu NSIT University of Chicago
Date: Mon, 26 Sep 2005 17:22:20 -0400 (EDT)
From: Alan Brown <alanb at digistar dot com>
Subject: Re: qpopper and quotas
On Mon, 26 Sep 2005, Edward Chase wrote:
> I'm having issues with my users blowing their storage quotas while checking
> their mail.
Set hard limit to 2.5* the soft limit.
From: "Alan W dot Rateliff, II" <lists at rateliff dot net>
Subject: RE: qpopper and quotas
Date: Tue, 27 Sep 2005 10:49:04 -0400
> -----Original Message-----
> From: Edward Chase [mailto:echase@studentweb.providence.edu]
> Sent: Monday, September 26, 2005 3:04 PM
> To: Subscribers of Qpopper
> Subject: qpopper and quotas
>
> I'm having issues with my users blowing their storage quotas
> while checking
> their mail.
> Anyone run into this and figure out a way around it?
On a Solaris system, I have the hard limit set at a reasonable amount about
the soft limit (reasonable based upon avergage usage) for /var/mail, and
have QPopper spool to a different filesystem.
The trick is that the quota system doesn't stop the user from spilling over
the hard limit until a certain time has expired. I changed the expiration
time on /var/mail to a few minutes. Even on a slow machine this gives
QPopper plenty of time to un-spool back to the mail store.
I think I pondered this on this list before, and this is the system I have
been using since without any problem.
--
Alan W. Rateliff, II : RATELIFF.NET
Independent Technology Consultant : alan2@rateliff.net
(Office) 850/350-0260 : (Mobile) 850/559-0100
-------------------------------------------------------------
[System Administration][IT Consulting][Computer Sales/Repair]
From: "Edward Chase" <echase at studentweb dot providence dot edu>
Subject: RE: qpopper and quotas
Date: Wed, 28 Sep 2005 07:52:02 -0400
No, /var and /var/mail are separate filesystems.
It's a Solaris box running sendmail.
> -----Original Message-----
> From: Ken A [mailto:ka@pacific.net]
> Sent: Monday, September 26, 2005 3:29 PM
> To: Edward Chase
> Cc: Subscribers of Qpopper
> Subject: Re: qpopper and quotas
>
>
> Are /var/mail and /var/spool/poptemp on the same partition?
>
> Are you running procmail as LDA ?
> It respects quotas.
> 'mail' does not.
>
> Ken
>
>
> Edward Chase wrote:
> > I'm having issues with my users blowing their storage
> quotas while checking
> > their mail.
> >
> > Consider this...
> >
> > User is at storage limit.
> > User checks mail.
> > /var/mail/user gets copied to /var/spool/poptemp/.user.pop
> > New mail comes into /var/mail/user
> > System cannot combine the 2 files back because it exceeds
> storage limit for
> > user.
> >
> > Obviously user is leaving mail behind on server...
> >
> > I haven't figured out a way around this.
> >
> > Anyone run into this and figure out a way around it?
> >
> > Thanks.
> >
>
From: "Edward Chase" <echase at studentweb dot providence dot edu>
Subject: RE: qpopper and quotas
Date: Wed, 28 Sep 2005 07:52:02 -0400
> At 03:04 PM 9/26/2005, Edward Chase wrote:
> >I'm having issues with my users blowing their storage quotas
> while checking
> >their mail.
> >
> >Consider this...
> >
> >User is at storage limit.
> >User checks mail.
> >/var/mail/user gets copied to /var/spool/poptemp/.user.pop
> >New mail comes into /var/mail/user
> >System cannot combine the 2 files back because it exceeds
> storage limit for
> >user.
> >
> >Obviously user is leaving mail behind on server...
> >
> >I haven't figured out a way around this.
> >
> >Anyone run into this and figure out a way around it?
>
> -----Original Message-----
> From: Daniel Senie [mailto:dts@senie.com]
> Sent: Monday, September 26, 2005 3:31 PM
> To: Edward Chase; Subscribers of Qpopper
> Subject: Re: qpopper and quotas
>
> > The only solutions I've seen involve putting the temp storage on a
> separate machine.
>
> The real answer, IMO, and one I haven't chased yet is to have the
> SMTP server (sendmail, etc.) handle mailbox sizing issues, and bounce
> messages when people are over quota. It should be possible to
> integrate such without that much effort, but the sendmail/procmail
> universe seems to lack such (I'm certainly open to pointers if folks
> have solutions to this).
>
> Ultimately, I really don't want to use the file system quota
> mechanism to deal with limiting users' mailbox sizes. Perhaps quotas
> work for you (other than the obvious qpopper question you asked
> about) but they seem like the wrong tool for the job.
>
using file system quota's was the only way I could see doing it. Care to
share in a little more detail on how to get sendmail to do this?
Thanks!
Date: Wed, 28 Sep 2005 15:26:21 +0300
From: Spiros Ioannou <sivann at image dot ece dot ntua dot gr>
Subject: Re: qpopper and quotas
Yes this is an old problem I had reported back in 2003. The only way arround
is to edit source code manually to change the spool locking mechanism.
Check out more info on these threads:
http://www.mail-archive.com/qpopper@lists.pensive.org/msg04089.html
-Spiros
Edward Chase wrote:
>>At 03:04 PM 9/26/2005, Edward Chase wrote:
>>
>>>I'm having issues with my users blowing their storage quotas
>>
>>while checking
>>
>>>their mail.
>>>
>>>Consider this...
>>>
>>>User is at storage limit.
>>>User checks mail.
>>>/var/mail/user gets copied to /var/spool/poptemp/.user.pop
>>>New mail comes into /var/mail/user
>>>System cannot combine the 2 files back because it exceeds
> using file system quota's was the only way I could see doing it. Care to
> share in a little more detail on how to get sendmail to do this?
>
> Thanks!
--
Spiros Ioannou
Image, Video and Multimedia Systems Laboratory
National Technical University of Athens
School of Electrical & Computer Engineering
Computer Science Division
Tel: +30-2107722491, +30-6973903808
From: "Edward Chase" <echase at studentweb dot providence dot edu>
Subject: RE: qpopper and quotas
Date: Wed, 28 Sep 2005 08:13:39 -0400
> On Mon, Sep 26, 2005 at 03:04:09PM -0400, Edward Chase wrote:
> > I'm having issues with my users blowing their storage
> quotas while checking
> > their mail.
> >
> > Consider this...
> >
> > User is at storage limit.
> > User checks mail.
> > /var/mail/user gets copied to /var/spool/poptemp/.user.pop
> > New mail comes into /var/mail/user
> > System cannot combine the 2 files back because it exceeds
> storage limit for
> > user.
> >
> > Obviously user is leaving mail behind on server...
> >
> > I haven't figured out a way around this.
> >
> > Anyone run into this and figure out a way around it?
> -----Original Message-----
> From: Clifton Royston [mailto:cliftonr@lava.net]
> Sent: Monday, September 26, 2005 4:03 PM
> To: Edward Chase
> Cc: Subscribers of Qpopper
> Subject: Re: qpopper and quotas
>
> >
> Yes, common known issue which admins have been batting around for
> years, with no entirely satisfactory solution.
>
> Do you currently have /var/spool/poptemp on the same file system as
> /var/mail? That exacerbates it.
>
> Generally AFAIK the best you can do is:
>
> * Run qpopper in server mode which optimizes away some spool copy
> operations;
>
> * Put /var/spool/poptemp on a separate FS;
>
> * Keep hard quota set to at least 2x soft quota (so that they can
> briefly spike to 2x the normal quota;
>
> * Set up a separate script which runs nightly and informs the users
> they are near to going over their (soft) quota, and to clean up
> their act!
>
> -- Clifton
Perhaps I can change qpopper to run in server mode.
I did try a soft quota of 2x the hard, but that doesn't work either.
Mail
will just keep coming in until it reaches the upper limit of the soft
quota.
At some point they are going to need more space just to do any cleanup.
Maybe if I try the soft quota along with some good scripting... Care to
share?
Thanks!
Date: Wed, 28 Sep 2005 08:47:40 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: qpopper and quotas
On Wed, Sep 28, 2005 at 08:13:39AM -0400, Edward Chase wrote:
...
Yes, server mode will help some because under a couple sets of
circumstances it eliminates copying/rewriting the spool in /var/mail.
(It also helps performance.)
> I did try a soft quota of 2x the hard, but that doesn't work either. Mail
> will just keep coming in until it reaches the upper limit of the soft quota.
> At some point they are going to need more space just to do any cleanup.
You mean hard quota = 2* the soft quota, right?
Yes, this does not solve the problem. It merely gives you a
high-water mark (and actually as somebody else noted you need the hard
quota to be 2*soft + some X for margin.) If you let it grow past the
soft quota, it will end up jamming once it gets > 1/2 the hard quota.
Using an LDA which can set limits on mailbox size is a better
solution. The Postfix LDA can do this if you want the same quota on
all mailboxes; I don't know what the sendmail method is.
> Maybe if I try the soft quota along with some good scripting... Care to
> share?
'Fraid I don't have anything on hand that I can turn over. We are
running some conglomeration of scripts to do that, but I'm not working
in that group (LavaNet Systems dept) any more, and I'm leaving this job
in a couple days.
-- Clifton
--
Clifton Royston -- cliftonr@tikitechnologies.com
Tiki Technologies Lead Programmer/Software Architect
"My own personal theory is that this is the very dawn of the world.
We're hardly more than an eyeblink away from the fall of Troy, and
scarcely an interglaciation removed from the Altamira cave painters. We
live in extremely interesting ancient times.
I like this idea. It encourages us to be earnest and ingenious and
brave, as befits ancestral peoples; but keeps us from deciding that
because we don't know all the answers, they must be unknowable and thus
unprofitable to pursue." -- Teresa Nielsen Hayden, 1995
Date: Wed, 28 Sep 2005 14:14:39 -0700
From: Ken A <ka at pacific dot net>
Subject: Re: qpopper and quotas
Some snippets for notifications (incomplete), but it should give you the
general idea. Combined with an 'educational warning message' to users,
and pruning repeat offender mailboxes, this works effectively.
Ken A.
use File::Find;
# $sentnoticefile nuked weekly via cron
# echo "">/usr/local/bin/sent_notice.dat
# so we don't send more than 1 notice per week to a user
open(SENT,"</$sentnoticefile");
@sent=<SENT>;
close (SENT);
$trigger_size = '50000000';
@exceptions = qw(root me etc); # leave these alone
chdir("/var/spool/mail");
find(\&biggie, ".");
sub biggie {
my $username = $_;
foreach $exuser(@exceptions) {
return if($exuser eq $username);
}
foreach (@sent) { chomp(); return if($_ eq $username); }
return unless -f && -s $username > $trigger_size;
# send the notice here via sendmail or whatever....
(I cut this part out for simplicity)
# then add the user to the sentnoticefile.
open(SENT,">>$sentnoticefile");
print SENT "$username\n";
close (SENT);
}
Edward Chase wrote:
>> On Mon, Sep 26, 2005 at 03:04:09PM -0400, Edward Chase wrote:
>>> I'm having issues with my users blowing their storage
>> quotas while checking
>>> their mail.
>>>
>>> Consider this...
>>>
>>> User is at storage limit.
>>> User checks mail.
>>> /var/mail/user gets copied to /var/spool/poptemp/.user.pop
>>> New mail comes into /var/mail/user
>>> System cannot combine the 2 files back because it exceeds
>> storage limit for
>>> user.
>>>
>>> Obviously user is leaving mail behind on server...
>>>
>>> I haven't figured out a way around this.
>>>
>>> Anyone run into this and figure out a way around it?
>
>
>> -----Original Message-----
>> From: Clifton Royston [mailto:cliftonr@lava.net]
>> Sent: Monday, September 26, 2005 4:03 PM
>> To: Edward Chase
>> Cc: Subscribers of Qpopper
>> Subject: Re: qpopper and quotas
>>
>> Yes, common known issue which admins have been batting around for
>> years, with no entirely satisfactory solution.
>>
>> Do you currently have /var/spool/poptemp on the same file system as
>> /var/mail? That exacerbates it.
>>
>> Generally AFAIK the best you can do is:
>>
>> * Run qpopper in server mode which optimizes away some spool copy
>> operations;
>>
>> * Put /var/spool/poptemp on a separate FS;
>>
>> * Keep hard quota set to at least 2x soft quota (so that they can
>> briefly spike to 2x the normal quota;
>>
>> * Set up a separate script which runs nightly and informs the users
>> they are near to going over their (soft) quota, and to clean up
>> their act!
>>
>> -- Clifton
>
>
> Perhaps I can change qpopper to run in server mode.
>
> I did try a soft quota of 2x the hard, but that doesn't work either. Mail
> will just keep coming in until it reaches the upper limit of the soft quota.
> At some point they are going to need more space just to do any cleanup.
>
> Maybe if I try the soft quota along with some good scripting... Care to
> share?
>
> Thanks!
>
>
>
>
Date: Wed, 28 Sep 2005 15:17:13 -0400 (EDT)
From: Alan Brown <alanb at digistar dot com>
Subject: RE: qpopper and quotas
On Wed, 28 Sep 2005, Edward Chase wrote:
> I did try a soft quota of 2x the hard, but that doesn't work either. Mail
> will just keep coming in until it reaches the upper limit of the soft quota.
The trick is to set the grace period between soft and hard quotas as
short as possible.
Date: Thu, 29 Sep 2005 14:05:20 +0300
From: Spiros Ioannou <sivann at image dot ece dot ntua dot gr>
Subject: Re: qpopper and quotas - a solution
The only way arround is to change the spool locking mechanism.
The spool (/var/mail/username) should stay locked as long the tmp
file exists. Right now it gets unlocked after the tmp file is created to allow
new mail to arrive, but if quota is exceeded we don't want new mail because
it will be appended to the old mail! The local delivery agent should wait if
it tries to deliver to a locked spool.
for an thourough explanation see
http://www.mail-archive.com/qpopper@lists.pensive.org/msg04089.html
and an answer from Randall Gellens:
http://www.mail-archive.com/qpopper@lists.pensive.org/msg04090.html
which sais that
"An option to hold the lock may make sense for some circumstances...this
should not be default behavior"
anyway we have modified the sources, and we have no more problems with quotas.
I've put up changes in source code for qpopper 4.0.5 (that could
be imported in newer versions). You can find the changes along with the
originals here:
http://manolito.image.ece.ntua.gr/~sivann/test/popper-ntua.tar.gz
Alan Brown wrote:
> On Wed, 28 Sep 2005, Edward Chase wrote:
>
>
>>I did try a soft quota of 2x the hard, but that doesn't work either. Mail
>>will just keep coming in until it reaches the upper limit of the soft quota.
>
>
> The trick is to set the grace period between soft and hard quotas as
> short as possible.
--
Spiros Ioannou
Image, Video and Multimedia Systems Laboratory
National Technical University of Athens
School of Electrical & Computer Engineering
Computer Science Division
Tel: +30-2107722491, +30-6973903808
Date: Thu, 29 Sep 2005 18:06:47 -0400 (EDT)
From: Edward Chase <echase at studentweb dot providence dot edu>
Subject: OT - imapd (was qpopper and quotas)
I just had a thought...
Would I be better off if my webmail users were using IMAP instead of POP?
Right now my webmail users are using Endymion Mailman.
http://www.endymion.com/products/mailman/
Would I be better off if they were using Squirrelmail?
http://www.squirrelmail.org/
Date: Thu, 29 Sep 2005 13:52:28 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: OT - imapd (was qpopper and quotas)
On Thu, Sep 29, 2005 at 06:06:47PM -0400, Edward Chase wrote:
> I just had a thought...
>
> Would I be better off if my webmail users were using IMAP instead of POP?
>
> Right now my webmail users are using Endymion Mailman.
>
> http://www.endymion.com/products/mailman/
In general IMAP is a better design fit with webmail applications...
> Would I be better off if they were using Squirrelmail?
>
> http://www.squirrelmail.org/
... however, IMAP can thrash your mailserver just as bad as popper,
if you don't use persistent connections in the webmail server, as it
will essentially run a fresh connection (or several) for every mouse
click.
This used to be a very difficult problem to solve, but in the past
couple years a session-caching imap proxy has gotten pretty solid. If
you configure Squirrelmail to connect to the port of the imap proxy,
the proxy to connect to your IMAP server, then your webmail server
should have pretty similar functionality to a regular mail client.
Disclaimer: I had this running and looking good, but never got it
into production before I was pulled off the project.
-- Clifton
--
Clifton Royston -- cliftonr@tikitechnologies.com
Tiki Technologies Lead Programmer/Software Architect
"My own personal theory is that this is the very dawn of the world.
We're hardly more than an eyeblink away from the fall of Troy, and
scarcely an interglaciation removed from the Altamira cave painters. We
live in extremely interesting ancient times.
I like this idea. It encourages us to be earnest and ingenious and
brave, as befits ancestral peoples; but keeps us from deciding that
because we don't know all the answers, they must be unknowable and thus
unprofitable to pursue." -- Teresa Nielsen Hayden, 1995
Date: Tue, 4 Oct 2005 07:06:44 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: qpopper and quotas - a solution
On Mon, Oct 03, 2005 at 05:11:02PM +0300, Spiros Ioannou wrote:
>
>
> Clifton Royston wrote:
> > This is not a bad idea; however, you should be aware that AFAIK
> >never-documented (but customary) behavior for mail delivery allows the
> >dot-lock to be broken after 5 minutes or some similar interval. This
> >could happen if a user is popping a large mailspool over a dialup
> >connection. Assuming you patched carefully, this won't worsen
> > problems, because qpopper can already deal with this; it just means
>
> yes, we take care of that with an alarm.. just search for the string "ntua"
> for the new functionality.
>
> >that in the quota-reached case you could still have the same result.
>
> why? the spool file will not grow while the tmp file is been processed.
> when the quota is been reached, no more mail will come to the spool and the
> user can successfully delete mail using pop.
I meant if you did not deal with the lock expiring; as it sounds like
you do, via stroking the lock on an alarm, then it's all good.
You would just need to be sure when using this that the mail server
LDA handles temporary delivery failures due to timeouts correctly, by
requeueing it for a later attempt on that mailbox.
-- Clifton
--
Clifton Royston -- cliftonr@iandicomputing.com / cliftonr@lava.net
I and I Computing - President
Custom programming, network design, systems and network consulting services
Date: Mon, 03 Oct 2005 17:11:02 +0300
From: Spiros Ioannou <sivann at image dot ece dot ntua dot gr>
Subject: Re: qpopper and quotas - a solution
Clifton Royston wrote:
> This is not a bad idea; however, you should be aware that AFAIK
> never-documented (but customary) behavior for mail delivery allows the
> dot-lock to be broken after 5 minutes or some similar interval. This
> could happen if a user is popping a large mailspool over a dialup
> connection. Assuming you patched carefully, this won't worsen
> problems, because qpopper can already deal with this; it just means
yes, we take care of that with an alarm.. just search for the string "ntua"
for the new functionality.
> that in the quota-reached case you could still have the same result.
why? the spool file will not grow while the tmp file is been processed.
when the quota is been reached, no more mail will come to the spool and the
user can successfully delete mail using pop.
With the old functionality when quota was exceeded and the spool was copied on
the tmp file, new mail could arrive and the total mail could be over the
quota; copying then the tmp file back to the spool would fail, leaving all the
old mail on the tmp file, and to the spool the new mail with some of the old
that could fit on the quotas.
-S
--
Spiros Ioannou
Image, Video and Multimedia Systems Laboratory
National Technical University of Athens
School of Electrical & Computer Engineering
Computer Science Division
Tel: +30-2107722491, +30-6973903808
Date: Thu, 13 Oct 2005 17:38:24 -0700 (PDT)
From: Kai Lanz <lanz at pangea dot Stanford dot EDU>
Subject: Broken pipe and timeout errors
We've just switched over to the new Qpopper v4.0.8; we were previously
using v2.53. Our server is an Alpha platform running Tru64 v4.0G. In the
log file, I now notice lots of "Broken pipe" messages:
Oct 13 16:48:55 geo-host popper[22043]: I/O error flushing output to client
joeuser at my-mac-g5.Stanford.EDU [123.45.67.89]: Broken pipe (32)
We get these about once a minute. Other messages logged by the same popper
process, like "Servicing request", "POP login", and "Stats", look okay.
We also see timeout notices, though much less frequently:
Oct 13 14:28:56 pangea popper[4268]: (v4.0.8-krb-V) Timeout (120 secs)
during nw read from johndoe at pc17.yale.edu (98.765.43.21)
What do these messages mean? Can we fix them? Can we ignore them?
From: "Edward Chase" <echase at studentweb dot providence dot edu>
Subject: RE: Broken pipe and timeout errors
Date: Fri, 14 Oct 2005 12:53:17 -0400
I believe you can ignore them...
The messages are caused by misbehaving client programs. (Timeouts too low
on the client side, I believe.)
> -----Original Message-----
> From: Kai Lanz [mailto:lanz@pangea.Stanford.EDU]
> Sent: Thursday, October 13, 2005 8:38 PM
> To: Subscribers of Qpopper
> Subject: Broken pipe and timeout errors
>
>
> We've just switched over to the new Qpopper v4.0.8; we were previously
> using v2.53. Our server is an Alpha platform running Tru64
> v4.0G. In the
> log file, I now notice lots of "Broken pipe" messages:
>
> Oct 13 16:48:55 geo-host popper[22043]: I/O error flushing
> output to client
> joeuser at my-mac-g5.Stanford.EDU [123.45.67.89]: Broken pipe (32)
>
> We get these about once a minute. Other messages logged by
> the same popper
> process, like "Servicing request", "POP login", and "Stats",
> look okay.
>
> We also see timeout notices, though much less frequently:
>
> Oct 13 14:28:56 pangea popper[4268]: (v4.0.8-krb-V) Timeout
> (120 secs)
> during nw read from johndoe at pc17.yale.edu (98.765.43.21)
>
> What do these messages mean? Can we fix them? Can we ignore them?
>
Date: Fri, 14 Oct 2005 14:31:42 -0400 (EDT)
From: Alan Brown <alanb at digistar dot com>
Subject: RE: Broken pipe and timeout errors
On Fri, 14 Oct 2005, Edward Chase wrote:
> The messages are caused by misbehaving client programs. (Timeouts too low
> on the client side, I believe.)
Correct.
If the volume is excessive some cluebatting of the user would be required.
AB
Date: Sun, 16 Oct 2005 23:25:04 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: "Password supplied is incorrect" error with Solaris 9...
At 12:48 PM -0500 9/26/05, Jerry K wrote:
> This is my configure string that I use for my solaris 9 systems. I
> came up with this based on other hints that I have read on this
> list, plus reading the docs. The enable-temp-drop-dir works well
> for me as I use user quotas.
>
> I am using the same version of gcc that you are.
>
> Hope this helps,
>
> Jerry Kemp
>
>
> ./configure --disable-apop \
> --enable-specialauth \
> --enable-uw-kludge \
> --disable-check-pwmax \
> --enable-log-login \
> --without-pam \
> --enable-temp-drop-dir=/var/tmp
It may be the "--without-pam" that is the key. The ./configure
script tries to figure out by itself if if '--enable-specialauth' is
needed, so normally you don't need to set it. If you run into a case
where ./configure does the wrong thing, let's try and fix it.
> Doug Wellington wrote:
>> I'm not able to login to Qpopper 4.0.8 running on a Solaris 9 server. I
>> keep getting the error:
>>
>> -ERR [AUTH] Password supplied for "ddw" is incorrect
>>
>> I can telnet and ftp to the server, but qpopper won't let me in. I looked
>> in the FAQ and did a google search, but the only thing I've found so far is
>> to use --enable-specialauth. I have done a make realclean and configured
>> both with and without --enable-specialauth and the binary ends up being
>> exactly the same. Is there an issue with which version of gcc I compile
>> with? (I'm using 3.4.2...)
I'd suggest enabling debug logging and see what gets logged. That
should give a hint as to what is wrong.
To enable tracing in Qpopper:
1. Do a 'make clean'
2. Re-run ./configure, adding '--enable-debugging'.
3. Edit the inetd.conf line for Qpopper, adding '-d' or '-t <tracefile-path>'.
4. Send inetd (or xinetd) a HUP signal.
Steps 3 and 4 are only needed if you use inetd (or xinetd); in
standalone mode, you can add '-d' or '-t <tracefile-path>' to the
command line directly.
Also, in either standalone or inetd mode, if you use a configuration
file you can add 'set debug' or 'set tracefile = <tracefile>' to
either a global or user-specific configuration file instead of steps
3 and 4.
This causes detailed tracing to be written to the syslog or to the
file specified as 'tracefile'.
Date: Fri, 18 Nov 2005 14:55:46 -0600
From: Tim Tyler <tyler at beloit dot edu>
Subject: getting poppassd on fedora 3 to work?
--=====================_21620562==.ALT
Content-Type: text/plain; charset="us-ascii"; format=flowed
Qualcomm popper and poppassd experts,
I recently downloaded 4.08 of qpopper to install on a Fedora core 3
server. I compiled qpopper along with poppassd using the special_auth
parameter. Correct me if I shouldn't include specail _auth. Popper
appears to work just fine. However, get the following error from our
webmail pop client from poppassd:
[]
Failed to change Password {400 Server busy -- try again later.}
I know that it sees the service, but the service is not responding
appropriately.
I put these entries in /etc/xinetd.d/poppassd
service poppassd
{
port = 106
socket_type = stream
protocol = tcp
user = root
server = /usr/local/bin/poppassd
server_args = poppassd
wait = no
only_from = 127.0.0.1
instances = 4
disable = no
}
Is the above entry correct? Am I missing something? Can anyone
suggest what I might be doing wrong to get the busy error above?
Tim Tyler
Network Engineer - Beloit College
tyler@beloit.edu
--=====================_21620562==.ALT
Content-Type: text/html; charset="us-ascii"
<html>
<body>
Qualcomm popper and poppassd experts,<br>
I recently downloaded 4.08 of qpopper to install on a
Fedora core 3 server. I compiled qpopper along with poppassd using
the special_auth parameter. Correct me if I shouldn't include
specail _auth. Popper appears to work just fine. However,
get the following error from our webmail pop client from poppassd:
<br><br>
<div align="center"><img src="http://netreg2.beloit.edu/cgi-bin/nwimg/mail/beloit/attention.gif" alt="[]"> Failed
to change Password {400 Server busy -- try again later.}<br><br>
</div>
I know that it sees the service, but the service is not responding
appropriately.<br><br>
I put these entries in /etc/xinetd.d/poppassd<br><br>
service poppassd<br>
{<br>
port = 106<br>
socket_type = stream<br>
protocol = tcp<br>
user = root<br>
server = /usr/local/bin/poppassd<br>
server_args = poppassd<br>
wait = no<br>
only_from = 127.0.0.1<br>
instances = 4<br>
disable = no<br>
}<br>
Is the above entry correct? Am I missing
something? Can anyone suggest what I might be doing wrong to get
the busy error above?<br><br>
<br>
<x-sigsep><p></x-sigsep>
Tim Tyler<br>
Network Engineer - Beloit College<br>
tyler@beloit.edu</body>
</html>
--=====================_21620562==.ALT--
Date: Fri, 18 Nov 2005 16:09:19 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: getting poppassd on fedora 3 to work?
--=====================_411133618==.ALT
Content-Type: text/plain; charset="us-ascii"; format=flowed
Because the POP Password service (poppassd) does not use encryption,
I generally recommend against its use. Use a web-based password
change mechanism instead.
As to your issues with special_auth, does popper itself interact
successfully with user account authentication, or is it just poppassd
that you have trouble with?
At 03:55 PM 11/18/2005, you wrote:
> Qualcomm popper and poppassd experts,
> I recently downloaded 4.08 of qpopper to install on a Fedora
> core 3 server. I compiled qpopper along with poppassd using the
> special_auth parameter. Correct me if I shouldn't include specail
> _auth. Popper appears to work just fine. However, get the
> following error from our webmail pop client from poppassd:
>
> Failed to change Password {400 Server busy -- try again later.}
>
> I know that it sees the service, but the service is not
> responding appropriately.
>
> I put these entries in /etc/xinetd.d/poppassd
>
>service poppassd
>{
>port = 106
>socket_type = stream
>protocol = tcp
>user = root
>server = /usr/local/bin/poppassd
>server_args = poppassd
>wait = no
>only_from = 127.0.0.1
>instances = 4
>disable = no
>}
> Is the above entry correct? Am I missing something? Can anyone
> suggest what I might be doing wrong to get the busy error above?
>
>
>Tim Tyler
>Network Engineer - Beloit College
>tyler@beloit.edu
--=====================_411133618==.ALT
Content-Type: text/html; charset="us-ascii"
<html>
<body>
Because the POP Password service (poppassd) does not use encryption, I
generally recommend against its use. Use a web-based password change
mechanism instead.<br><br>
As to your issues with special_auth, does popper itself interact
successfully with user account authentication, or is it just poppassd
that you have trouble with?<br><br>
At 03:55 PM 11/18/2005, you wrote:<br>
<blockquote type=cite class=cite cite=""> Qualcomm popper and
poppassd experts,<br>
I recently downloaded 4.08 of qpopper to install on a
Fedora core 3 server. I compiled qpopper along with poppassd using
the special_auth parameter. Correct me if I shouldn't include
specail _auth. Popper appears to work just fine. However, get
the following error from our webmail pop client from poppassd:
<br><br>
<div align="center"> Failed to change Password {400 Server busy --
try again later.}<br><br>
</div>
I know that it sees the service, but the service is not responding
appropriately.<br><br>
I put these entries in /etc/xinetd.d/poppassd<br><br>
service poppassd<br>
{<br>
port = 106<br>
socket_type = stream<br>
protocol = tcp<br>
user = root<br>
server = /usr/local/bin/poppassd<br>
server_args = poppassd<br>
wait = no<br>
only_from = 127.0.0.1<br>
instances = 4<br>
disable = no<br>
}<br>
Is the above entry correct? Am I missing
something? Can anyone suggest what I might be doing wrong to get
the busy error above?<br><br>
<br>
Tim Tyler<br>
Network Engineer - Beloit College<br>
tyler@beloit.edu </blockquote></body>
</html>
--=====================_411133618==.ALT--
Last updated on 18 Nov 2005 by Pensive Mailing List Admin