ALERT: exim paniclog /var/log/exim/paniclog has non-zero size, mail system possibly b

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
Hi,

I don´t know if it is related but after installing new evolution skin I get a new message in my inbox:
ALERT: exim paniclog /var/log/exim/paniclog has non-zero size, mail system possibly broken

Never saw this alert before. Is there a fix/solution for this ?

Kind regards,
Fred
 

zEitEr

Super Moderator
Joined
Apr 11, 2005
Messages
14,094
Location
GMT +7.00
Hello Fred,

And what do you have in /var/log/exim/paniclog?

If you are on FreeBSD ... you might ignore that alert about /var/log/exim/paniclog as it gets a rotation notice in the file.


 

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
Hi Alex,

It is on Centos 7.

For now I just removed everything inside the file. I read something about that as a solution.
Now let´s wait if the message returns (after a reboot?).

Kind regards, Fred
 

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
This night I receive the same alert form Directadmin.

The exim /var/log/exim/paniclog has a non-zero size, which usually means something in exim is very broken. Please review the last 20 lines, below, and address the issue.
This message will not resend for 24 hours.
In this log I see:
2018-01-06 11:52:24 1eXm5Q-0004Fq-JF unable to open private key file for reading: /etc/virtual/server.myservername.com/dkim.private.key
I seems that I need to have a dkim key for my hostname? apparently this
echo "action=rewrite&value=dkim" >> /usr/local/directadmin/data/task.queue
is not creating a dkim for the hostname??

Anyway, I did for now:
cd /usr/local/directadmin/scripts
./dkim_create.sh server.myservername.com
If this solve the alert, I will let you know.

Kind regards, Fred
 

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
Unfortunately my "solution" did not work out. Same message again. Anyone an idea?
 

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
Hi Alex,

Same message. Maybe I have to remove it again from paniclog, as it seems that it is not a new alert and as it is not empty, I receive the alert. First time I removed everything from Paniclog, but after creating dkim key for hostname, I didn´t.
 

zEitEr

Super Moderator
Joined
Apr 11, 2005
Messages
14,094
Location
GMT +7.00
The logs by default are rotated weekly, so you need either to force logrotation or to empty it manually.
 

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
Thanks Alex, I will do and let you know if the problem is solved.

Kind regards, Fred

PROBLEM SOLVED, THANKS AGAIN
 
Last edited:

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
Hi,

As I mentioned in my first question, all alerts started after installing the new Evolution skin. I solved one, but now I receive other alerts in Paniclog:

018-01-09 11:03:14 1eYqkM-0002aq-2R Failed writing transport results to pipe: Broken pipe
2018-01-09 11:03:14 1eYqkM-0002aq-2R Failed writing transport results to pipe: Broken pipe
2018-01-09 11:03:14 1eYqkM-0002aq-2R Failed writing transport results to pipe: Broken pipe
2018-01-09 19:43:28 1eYyrk-0004va-Nb malware acl condition: clamd [127.0.0.1]:3310 : unable to read from socket (Connection timed out)
Before I never received any alert related to Paniclog.

Any ideas?
 

zEitEr

Super Moderator
Joined
Apr 11, 2005
Messages
14,094
Location
GMT +7.00
Re-configure exim not to scan emails with ClamAV, or make sure ClamAV is running.
 

Awd

Verified User
Joined
Aug 9, 2015
Messages
316
Hi Alex,

ClamAV is running. After kill all processes I rebuyld again. Hopefully it will fix it.
One other thing I remember, almost at the same time I installed new skin, I did also some security kernel patches.

Fact is that some alerts since a couple of years suddenly appears.
 

zEitEr

Super Moderator
Joined
Apr 11, 2005
Messages
14,094
Location
GMT +7.00
A new skin hardly can be a root cause of the issue. So you need to check history of changes. Probably you changed something that causes the error:

Code:
malware acl condition: clamd [127.0.0.1]:3310 : unable to read from socket (Connection timed out)
or clamav was too busy to process a new request, or it was in a process of restarting. So you should check more logs for more details.
 
Top