Exim 4.95

CustomBuild 2.0 rev. 2369 with steroids

Hello, CB 2.0 rev. 2369 has been released. It's a huge release for developers :) As it now lets you overwrite any call, execute pre/post hooks etc. For example, you could now: 1) create "./build varnish" or just overwrite regular "./build apache" call with whatever you want by using pre-call...
forum.directadmin.com forum.directadmin.com

Works great. I use these hooks for example to send Pushover notifications when there are updates available.
 
wattie said:
After updating to this version the Exim stops working with the following errors in /var/log/messages:

Code: 
Nov 29 12:32:44 srv2 kernel: pid 23374 (exim), jid 0, uid 12: exited on signal 11
Nov 29 12:32:48 srv2 kernel: pid 23483 (exim), jid 0, uid 12: exited on signal 11
Nov 29 12:32:49 srv2 kernel: pid 23784 (exim), jid 0, uid 12: exited on signal 11

The exim panic log shows messages like these:

Code: 
2021-11-29 12:32:44 1mrdxi-000650-QY SIGSEGV (maybe attempt to write to immutable memory)
2021-11-29 12:32:47 1mrdxm-00066l-Nu SIGSEGV (maybe attempt to write to immutable memory)
2021-11-29 12:32:49 1mrdxo-0006Bc-Ps SIGSEGV (maybe attempt to write to immutable memory)

I had to revert back to 4.94.2
Click to expand...
On FreeBSD disable TCP socket in ClamAV (let only local socket to work) and modify exim.clamav.load.conf (exim.clamav.load.conf.custom ?) accordingly.
 
shanti said:
@sparek i would like to know more about this way to trigger hook-scripts .. how you came to that info ? would pls you be so kind and tell me where you found info about this ?

tnx
-c-
Click to expand...
How I came to know about it? No clue. Can't remember.

Probably in a forum post some where.

That's kind of the issue I have with DirectAdmin sometimes. The documentation leaves, well, really a lot be desired.

And the documentation that does exist, often times seems to assume that you played a role in designing or developing the stated item so it's not always heavy on details.

That's probably one of the reason this forum has a lot of "how do I....?" posts (and also because people prefer to be spoon fed information rather than actually looking for it). There might be documentation some where on the DirectAdmin site that explains the "how do I...?" question, but finding it, and then understanding it just leaves a lot to be desired.
 
sparek said:
There might be documentation some where on the DirectAdmin site that explains the "how do I...?" question, but finding it, and then understanding it just leaves a lot to be desired.
Click to expand...
Yup and that costs Support time and tickets also i guess. ?

Circle then less time to spend on the DOCS.
 
I'm encountering warnings like this in the log.
I have the exim.conf 4.5.36 with the new maximun line length setting and now I see this in an incoming mail in the Exim logfile.
Code: 
to <[email protected]> exceeds RFC 5322 maximum line length (1149 > 998)
I did get the e-mail as this was only present as a warning in the log, nothing critical. But shouldn't the warnings be gone too?
 
Last edited:
Richard G said:
I have the exim.conf 4.5.36 with the new maximun line length setting and now I see this in an incoming mail in the Exim logfile.
Code: 
to <[email protected]> exceeds RFC 5322 maximum line length (1149 > 998)
Click to expand...
If you added the lines from @kam821 in post 62 - (see the bottom of that post) if you added it into the /etc/exim.acl_check_message.pre.conf file as suggested, then yes, you'll still get the "warning" message in your logs. But it doesn't disrupt anything, it's just a "warning" (i.e. a way to add stuff into the log file).
 
sparek said:
That's kind of the issue I have with DirectAdmin sometimes. The documentation leaves, well, really a lot be desired.

And the documentation that does exist, often times seems to assume that you played a role in designing or developing the stated item so it's not always heavy on details.
Click to expand...
That's why I created this thread: Knowledge Base - Errors or omissions, got confirmation by a DirectAdmin employee they would take input from there to update their docs pages. Maybe it helps?
 
sirwiz said:
After update to 4.95 SMTP stop working, from log I got:
failed to open /etc/virtual/whitelist_hosts_ip when checking "/etc/virtual/whitelist_hosts_ip": Permission denied (euid=8 egid=12)
And also few other files.
Sure, I fix it, now it working but strange is that until update it works without any problems.
(to fix it just go to /etc/virtual, and chown mail:mail nesesery files)
Click to expand...
Same here after updating Exim to 4.95
 
Imtek said:
Same here permission issues without changing anything only the available/installed newer Exim version, any idea @smtalk @DirectAdmin Support ?
Click to expand...
@Remco00 @Imtek
It looks, like there may be some issue in Directadmin binary/scripts, because this file has different chmod value than other whitelist_ files.
-rw-r--r-- 1 mail mail 0 08-11 13:50 whitelist_domains
-rw-r--r-- 1 mail mail 0 08-11 13:50 whitelist_from
-rw-r--r-- 1 mail mail 0 08-11 13:50 whitelist_hosts
-rw------- 1 mail mail 0 08-11 13:36 whitelist_hosts_ip
-rw-r--r-- 1 mail mail 0 08-11 13:50 whitelist_senders

Try to execute:
chmod 644 /etc/virtual/whitelist_hosts_ip

Overall, something weird is going on with /etc/virtual files permissions.
They are are not consistent and I don't think that in many cases exposing content to the public is necessary.
 
Last edited:
set_permissions.sh and CustomBuild sets chmod 600 and mail:mail ownership there. May you try running set_permissions.sh to fix them up?
 
+1 for set_permissions.sh all. Always use it after DA / System updates and never have any permission issues.
 
Hi,

I know I'm late, but I would like someone's help with my situation, please.

Are there cases where the way the email software handles messages is solely responsible for causing this "message has lines too long to transport" problem? I have this question because I get this error even after setting the limit stated in post #58 to 999999999999

I noticed an unusual behavior in relation to the occurrence of this problem. Sometimes even the most up-to-date versions of email software used by customers get this error, but what all these occurrences have in common is a certain number of replies/interactions in the same message. In some cases even in RoundCube this error occurs if the message has a certain number of responses/interactions.

In these cases, is there any way to validate, such as parsing the full content of the message, for example, to determine which specific content of the message is causing the problem?

I don't know, but it could be that the recipients' servers are refusing messages with this error... Are there any examples of e-mail contents that cause this error? It would be extremely valuable for analysis in specific cases.
 
Serg951 said:
I have this question because I get this error even after setting the limit stated in post #58 to 999999999999
Click to expand...
As you could see, this has already been fixed, so you should not edit directly the exim.conf anymore, hard to notice where it's going wrong.
Just updating the exim.conf should do the trick.
If all is well, you should have version 4.5.37 of exim.conf in the /etc directory.
Also, you should really be 100% sure that it is DA causing the issue and not a remote server.

Serg951 said:
but it could be that the recipients' servers are refusing messages with this error...
Click to expand...
I've seen this happening with cPanel indeed. cPanel keeps the limit at 2048, users who want this changed now have a seperate option to change this to a higher number in the Exim configuration editor. For example logwatch e-mails from cPanel will cause this issue for sure.
Which is the reason I had to change that value on cP also.
 
You must log in or register to reply here.
Back
Top