Mail system problems.

[SkIRmiS]

Hello,

1. first problem.
Sometimes I got the strange message:
with no subject, name and et.c the body is:

X-Spam-Checker-Version: SpamAssassin 3.0.4 (2005-06-05) on changed.com
X-Spam-Level:
X-Spam-Status: No, score=-1.2 required=5.0 tests=ALL_TRUSTED,MISSING_DATE,
MISSING_SUBJECT autolearn=unavailable version=3.0.4

I got this message after sending file.txt to my email. Already, i got 5 messages like this. My email is set to system admin email in ISPconfig.

2. second problem, with quotas.
Web users and web site has quota 5000 mb. The ISPconfig shows that site are only using 23.2 mb of all 5000. Web site users does'nt has mail box quota limits. But sometimes whey cannot receive email, of a error Mail quota exceeded!

On shell:
repquota -avug | grep web29
web29 -- 410904 5120000 5120000 6992 0 0

As you can see web29 user are using 401 mb. Not 23.2 mb that says the ISPconfig. I checked the owners of users directories, everything is fine. Others sites are showing not correct size too:( But at least the emails are working well.

3. Where to find logs of spammassins reports, trashscan, clamav?

4. Thanks for advice.

Skirmis

[till]

2. second problem, with quotas.
Web users and web site has quota 5000 mb. The ISPconfig shows that site are only using 23.2 mb of all 5000. Web site users does'nt has mail box quota limits. But sometimes whey cannot receive email, of a error Mail quota exceeded!

On shell:
repquota -avug | grep web29
web29 -- 410904 5120000 5120000 6992 0 0

As you can see web29 user are using 401 mb. Not 23.2 mb that says the ISPconfig. I checked the owners of users directories, everything is fine. Others sites are showing not correct size too:( But at least the emails are working well.

ISPConfig reports the size of the website directory including the subdirectorys user, logs, web, etc. If your users are able to store files in other places the sizes may differ.

3. Where to find logs of spammassins reports, trashscan, clamav?

They dont write logs by default.

[SkIRmiS]

ISPConfig reports the size of the website directory including the subdirectorys user, logs, web, etc. If your users are able to store files in other places the sizes may differ.


They dont write logs by default.

1.The users don't have other places outside webXX.
2.I need these logs to see that happens with rejected mail.

And do you know something about my first problem?


Thanks.

[jmroth]

Well well,
I have started receiving quota exceeded errors for one user too, and the user definitely has no quota restrictions (neither does the group).
The debug output shows interesting things:
/bin/sh: line 1: let: HD_QUOTA=*1024: syntax error: operand expected (error token is "*1024")
Apparently procmail thinks QUOTA is set and then tries to do arithmetics on an empty expression which obviously goes wrong.
I didn't dig into procmail yet but maybe this helps someone figure out the real problem.
Regards,
~jm

[jmroth]

I don't really know but maybe what you were trying to do should be
... quota -v ${LOGNAME} | sed -n 3p ....
on line 22 of .quota.rc, otherwise the 3p wouldn't really make sense
however I do not know why my other users don't generate error messages
very strange although something is obviously wrong here

[Tenaka]

Hello,

1. first problem.
Sometimes I got the strange message:
with no subject, name and et.c the body is:

X-Spam-Checker-Version: SpamAssassin 3.0.4 (2005-06-05) on changed.com
X-Spam-Level:
X-Spam-Status: No, score=-1.2 required=5.0 tests=ALL_TRUSTED,MISSING_DATE,
MISSING_SUBJECT autolearn=unavailable version=3.0.4


I have exactly the same strange, empty emails, I get them around 5-7 times a day, its hard to track them down as they don't even have a timestamp, nothing. I had one idea, concerning their origin: it seemed like when I was trying to send an email to a non-existent mailbox I got these empty mails instead of a user unknown reply. but I can't test it as I can't send emails to local domains, because these get instantly verified...

can you try and send an email to nonexistent@web-designerz.de ? Please check if you get one ofthese empty mails right after that.

[brent.stephens]

I'm getting these extremely odd mail quota issues now as well. I cannot discern why its happening, but its extremely frustrating and pretty embarassing too. syslog output below. It seems to be breaking up their 'real' name and trying to send it to them...I just have no idea where its getting it and WHY its happening...


Jan 2 13:07:46 liquid5th postfix/smtpd[955]: connect from edited[xx.xx.xx.xx]
Jan 2 13:07:46 liquid5th postfix/smtpd[955]: 715DB11033: client=edited[xx.xx.xx.xx], sasl_method=PLAIN, sasl_username=myuser
Jan 2 13:07:46 liquid5th postfix/cleanup[959]: 715DB11033: message-id=<4366AD72-E6E8-4E0C-8BD9-9B8D2CB8B54A@liquid5th.com>
Jan 2 13:07:46 liquid5th postfix/local[960]: 715DB11033: to=<davesperandio@hosting.liquid5th.com>, orig_to=<dave@allforacappella.org>, relay=local, delay=0, status=sent (delivered to command: /usr/bin/procmail -f-)
Jan 2 13:07:46 liquid5th postfix/pickup[31970]: D96F711033: uid=15011 from=<davesperandio>
Jan 2 13:07:46 liquid5th postfix/cleanup[959]: D96F711033: message-id=<20060102180746.D96F711033@liquid5th.net>
Jan 2 13:07:47 liquid5th postfix/local[960]: D96F711033: to=<admispconfig@localhost>, relay=local, delay=1, status=sent (delivered to command: /usr/bin/procmail -f-)
Jan 2 13:07:48 liquid5th postfix/pickup[31970]: 0A74511039: uid=15011 from=<davesperandio>
Jan 2 13:07:48 liquid5th postfix/cleanup[959]: 0A74511039: message-id=<20060102180747.0A74511039@liquid5th.net>
Jan 2 13:07:48 liquid5th postfix/pickup[31970]: 18B0A11033: uid=15011 from=<David>
Jan 2 13:07:48 liquid5th postfix/cleanup[959]: 18B0A11033: message-id=<20060102180748.18B0A11033@liquid5th.net>
Jan 2 13:07:48 liquid5th postfix/local[960]: 0A74511039: to=<admispconfig@localhost>, relay=local, delay=1, status=sent (delivered to command: /usr/bin/procmail -f-)
Jan 2 13:07:48 liquid5th postfix/local[1037]: 18B0A11033: to=<Sperandio@hosting.liquid5th.com>, orig_to=<Sperandio>, relay=local, delay=0, status=bounced (unknown user: "sperandio")
Jan 2 13:07:48 liquid5th postfix/pickup[31970]: 4241311315: uid=15011 from=<davesperandio>
Jan 2 13:07:48 liquid5th postfix/cleanup[1026]: 4241311315: message-id=<20060102180748.4241311315@liquid5th.net>
Jan 2 13:07:48 liquid5th postfix/local[1037]: 4241311315: to=<admispconfig@localhost>, relay=local, delay=0, status=sent (delivered to command: /usr/bin/procmail -f-)
Jan 2 13:07:50 liquid5th postfix/local[960]: 18B0A11033: to=<davesperandio@hosting.liquid5th.com>, orig_to=<dave@allforacappella.org>, relay=local, delay=2, status=sent (delivered to command: /usr/bin/procmail -f-)
Jan 2 13:07:50 liquid5th postfix/cleanup[959]: 302761103B: message-id=<20060102180750.302761103B@liquid5th.net>
Jan 2 13:07:50 liquid5th postfix/local[1037]: 302761103B: to=<David@hosting.liquid5th.com>, relay=local, delay=0, status=bounced (unknown user: "david")

[till]

I dont think this are quota errors. What is the linux username for example for the email address: David@hosting.liquid5th.com ?

[brent.stephens]

The username is davesperandio

[falko]

Is David@hosting.liquid5th.com in /etc/postfix/virtusertable? If so, to what user is it mapped?

[momo]

I have the same problem too:
for example:

#
cat virtusertable | grep audrone
audrone@changed.tld audrone
audrone@changed.tld audrone
#
repquota -avug | grep audrone
audrone -- 16468 0 0 113 0 0
#
/tmp/quota.audrone exist!

And the user cannot receive emails, got Mail Quota Exceeded...

[till]

You mean you have the same problem as skirmis, but not the same problem then brent.stephens.

Run:

rm -f /tmp/quota.*

I've posted a solution how to disable the quota script in this thread:

http://www.howtoforge.com/forums/showthread.php?t=1653

[momo]

You mean you have the same problem as skirmis, but not the same problem then brent.stephens.

Run:

rm -f /tmp/quota.*

I've posted a solution how to disable the quota script in this thread:

http://www.howtoforge.com/forums/showthread.php?t=1653


Yes, I mean this, I fallowed your advice:
If you use maildir, you can try to disable the quota procmail recipe. Then the quota is handled completely by the linux system quota. This might solve the problem.

Open the file

/root/ispconfig/isp/conf/procmailrc.master

and comment out the line:

INCLUDERC={PMDIR}/.quota.rc

Then go in the ispconfig interface and change something in an account of y user where you got this problem and save the user. The procmailrc will then be rewritten.

=============================

But don't know how to change to all users this ? Not to only 4 one... I have about 600 users

Thanks for advice!

[falko]

But don't know how to change to all users this ? Not to only 4 one... I have about 600 users

Thanks for advice!
You could either write a script that does it, or you set the status field of all users in the isp_isp_user table in the ISPConfig database to u and run touch /home/admispconfig/ispconfig/.run
This will make ISPConfig rewrite the procmail recipes (I haven't tested it, but it should work).

[rafaelenrike]

I had almost the same problem. this is the error email that returns back when I send an email to a specific account (not all accounts):

This is the Postfix program at host mailserver.domain.tld.

I'm sorry to have to inform you that your message could not
be delivered to one or more recipients. It's attached below.

For further assistance, please send mail to <postmaster>

If you do so, please include this problem report. You can
delete your own text from the attached returned message.

The Postfix program

<web1_user1@mailserver.domain.tld> (expanded from <user1@domain.tld>):
can't create user output file. Command output: /bin/sh: let:
HD_QUOTA=*1024: syntax error: operand expected (error token is "*1024")
/bin/sh: line 1: [: !=: unary operator expected procmail: Error while
writing to "/var/mail/web1_user1"

Does anyone know why is dis happening and how to fix it?

Thanks for your help.

[till]

Please set a web (harddisk) quota for this user in ISPCOnfig.

Do you use maildir or mbox format for mail storage?

[rafaelenrike]

Please set a web (harddisk) quota for this user in ISPCOnfig.

Do you use maildir or mbox format for mail storage?

I use mbox format for mail storage. Every account for this domain is using mbox format and also every account has set 0 Mb for Web storage and XX Mb for Mail storage.

[till]

I use mbox format for mail storage. Every account for this domain is using mbox format and also every account has set 0 Mb for Web storage and XX Mb for Mail storage.

Do you have linux system quota enabled on your /var partition?

If yes, you must set the webspace quota at least to the double size of the mbox size.

[rafaelenrike]

Do you have linux system quota enabled on your /var partition?

If yes, you must set the webspace quota at least to the double size of the mbox size.

I just have "/boot" and "/" partitions on my system (I used FC4 perfect setup). I already configured quotas for "/".

[root@webserver ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
287G 3.5G 269G 2% /
/dev/sda2 99M 23M 72M 24% /boot
/dev/shm 504M 0 504M 0% /dev/shm

So I just have one partition for all the system.

Thanks for your help.

[falko]

Please post the output of ls -la /var/mail and repquota -avug | web1_user1

What's web1_user1's homedir?

[rafaelenrike]

Please post the output of ls -la /var/mail and repquota -avug | web1_user1

What's web1_user1's homedir?

This is the output for ls -la /var/mail for the specified user (I can't post all the directory output because it's too long (so many characters).
[root@webserver ~]# ls -la /var/mail/web1_user1
-rw------- 1 web1_user1 mail 2910002 Apr 4 16:33 /var/mail/web1_user1

This is the output for repquota -avug | web1_user1

[root@webserver ~]# repquota -avug | grep web1_user1
web1_user1 -- 2864 0 0 5 0 0

The home directory is at /home/www/web1/user/ (the ISPConfig default).

[falko]

Looks ok. What's the output of df -h?

[fobicodam]

Looks ok. What's the output of df -h?

everything is fine in my server, it keeps telling me procmail has not rights to write in the mail dir, pleaseee

[till]

Can you please post the output of the command "df -h" as falko requested?

[fobicodam]

Will be usual to fix one problem to have another? im really tired of fix bugs, i dont know if they are linux or postfix or ispconfig's bugs, but im really tired..

How the hell can i fix this one?!?!?!

This is the Postfix program at host webserver2.imago.com.ar.

I'm sorry to have to inform you that your message could not
be delivered to one or more recipients. It's attached below.

For further assistance, please send mail to <postmaster>

If you do so, please include this problem report. You can
delete your own text from the attached returned message.

The Postfix program

<web1_daniel@webserver2.imago.com.ar> (expanded from <daniel@imago.com.ar>):
can't create user output file. Command output: procmail: Couldn't create
"/var/mail/web1_daniel"



The logs says the "·$)"·($") HD_QUOTA=*1024 error.

[till]

Which ISPConfig version do you use?
Do you use Maildir or Mbox as mail storage format?
Please post the output of the command "df -h" as falko requested?

If you dont post data like the output of the command "df -h", we are not able to help you!

[tijn_tux]

Well, correct me if iam wrong, but if you use mbox you can max have a mailfile of x mb (really not mucht) i've had the same problem, only solution i had was to split the whole mailfile into sinlge mails and use cat to restore the original, (meanwhile fetching all the mail)

Hope this helps you...

tijn

[till]

On current linux distributions, you dont have the 2GB limit anymore. But mbox files tend to get corrupted when a mail arrives while you download emails with pop3 or access them with imap. Thats why we recommend always to use maildir.