Daily problems with Exim

G

GuyNijssen

Verified User
Joined
Jul 1, 2020
Messages
9
Hi all

Since 2 weeks, since the activation of DKIM (not sure if it is related) I'm experiencing problems with Exim.
Every morning I see incoming and outgoing mails are "stuck" in the mail queue for 30 minutes or less, after which they get delivered.
I cannot find any reason in the mail logs, just the first step taking place:

2020-07-01 06:24:20 1jqUIA-00007H-Cw <= guy@***.be H=(***.be) [127.0.0.1] P=esmtp S=337 T="test 14" from <guy@***.be> for [email protected]

After a restart of exim (which takes minutes!) the mails start to come in normally, and future mails continue to flow in without any delay.
Future restarts of exim are a lot faster.

I'm not sure which version of exim was running at the time the problems started, but since 2 days I'm running on the latest version (4.94)
Exim 4.94
OS: Centos 7

I did a full update of Centos, followed by a full update and rebuild of Directadmin Custombuild, but this didn't fix the problem.

Does anybody has an idea?

Thanks in advance
Best regards,
Guy
 
Do you have Easy Spam Fighter installed ?
One of our servers had this problem also , after removing this and rebuilding email stuck problem was solved
I am not really sure if ESF was the problem but it helped
 
I removed all anti-spam software in the very beginning, because spamd was the first deamon that failed.
But it failed probably because of the exim problem...
After removing, I rebuilt all software...
 
In this example, I did a telnet to port 25 on the server to see what response i got (the answer is none):
After typing the dot, nothing happens... No reponse, and no logging.
Sending from an external mailserver, or the webmail results in the same problem.
 
Owz that's very weird. I wanted to suggest you rebuild exim and exim.conf but if I read correctly you already have done this.
I presume there is also no high cpu usage? Just Exim not responding.

You thought it might have started after installing DKIM. What nameservers are you using? External nameservers or local ones?
You hopefully do not use one of the Google servers like 8.8.8.8 in the /etc/resolv.conf file?
Which version of Exim and which version of exim.conf and exim.pl do you have?
 
Richard G said:
Owz that's very weird. I wanted to suggest you rebuild exim and exim.conf but if I read correctly you already have done this.
I presume there is also no high cpu usage? Just Exim not responding.
Click to expand...
No, no high cpu usage, and Exim is responding slow, too slow to send mail, but fast enough to receive mail with a delay of 30 to 60 minutes. One thing I didn't mention: When the problem occurs, mails come in multiple times.

Richard G said:
You thought it might have started after installing DKIM. What nameservers are you using? External nameservers or local ones?
You hopefully do not use one of the Google servers like 8.8.8.8 in the /etc/resolv.conf file?
Click to expand...
Actually, I am:
# Generated by NetworkManager
search 24it.be
nameserver 8.8.8.8
nameserver 8.8.4.4
nameserver 208.67.222.222
# NOTE: the libc resolver may not support more than 3 nameservers.
# The nameservers listed below may not be recognized.
nameserver 208.67.220.220
Click to expand...

Richard G said:
Which version of Exim and which version of exim.conf and exim.pl do you have?
Click to expand...
Exim: 4.94
Exim.pl: version=29
Exim.conf: How can I find out this version?

Already thank you for trying to help me!
 
Try reverting to exim 4.93.0.4 and look if this helps
 
Ok, did this:
cd /usr/local/directadmin/custombuild
echo "exim:4.93.0.4:" > custom_versions.txt
./build exim
./build exim_conf
Click to expand...
I'll keep you updated on the result tomorrow morning.
Thanks
 
floyd said:
What's wrong with Google's servers?
Click to expand...
In principle, nothing, but they limit the amount of lookups, good chance you will run into that at a certain time. I never had that issue and suddenly it started.
It's mentioned also by smtalk already on the forums somewhere. If you run your own nameservers you can best run 127.0.0.1 as nameserver, and if you want to use external nameservers you can prevent running into limits by using 1.1.1.1 or opendns as nameservers.

GuyNijssen said:
Actually, I am:
Click to expand...
Better change them to:
1.1.1.1
1.0.0.1
and then the others you have.
 
Richard G said:
In principle, nothing, but they limit the amount of lookups, good chance you will run into that at a certain time. I never had that issue and suddenly it started.
It's mentioned also by smtalk already on the forums somewhere. If you run your own nameservers you can best run 127.0.0.1 as nameserver, and if you want to use external nameservers you can prevent running into limits by using 1.1.1.1 or opendns as nameservers.
Click to expand...
Interesting. But as the problem seemed to be soved, I'll wait a few days before changing anything else.
 
Oke, nice to hear the older version does not cause issues.
Makes me wonder if some little thing went wrong on upgrading, or the newer Exim is causing issues some other way, but not to all systems.
 
Well I was having the problem on 2 servers, but not that regular on the 2nd server.
But the 2nd server is barely used (6 domains against 250 on the first).
Just saying, it didn't happen again. Thank you all for your support!
This is a great community!
 
Might be related to this thread:

Exim 4.94 - Eximconf 4.5.24 - CentOS7 - Exim problem after some minutes

We have a problem after installing Exim 4.94 (and eximconf 4.5.24) on a CentOS 7 server. (we have also CentOS 8 servers and this problem doesn't exists on these servers! ) The update from 4.93.0.4 to 4.94 is done correctly, without any error. After installing and restarting exim works good. But...
forum.directadmin.com forum.directadmin.com

Als some issues where on several systems, the only solution is to revert to the older version of Exim.

Glad you got it fixed now!
 
You must log in or register to reply here.
Back
Top