Outbound email issues for one domain

[thowden]

Hi all

I have two CPanel servers (well, 2 I am interested in today):
#1. Hosting server - at a host that blocks port 25 mail outbound.
#2. Hosting server - with a host that allows port 25 mail outbound

Server #1 uses -> Server #2 as a "smarthost"

#whmapi1 installed_versions packages=1|grep exim
#cpanel-exim: 4.94.2-4.cp1198
- cpanel-exim

Both servers consistent with exim (as above) and cpanel 98.0.8 versions. Differences? #1 is CentOS8.4 while #2 is CentOS7.8 - CloudLinux

Mail seems to flow for most / many of my hosted clients (sample testing shows no other unreasonable errors), so almost all good.

One specific domain sender gives me errors sending to an external client self-hosted server (SBS2011 - Exchange), but seems to send ok otherwise.

The specific messages bounce with failure message

"all hosts for 'externalserver.sbs2011.domain' have been failing for a long time (and retry time not reached)"

or

"retry timeout exceeded"

and initial defer messages

"Connection timed out H=my.server#2 [ip.address]: SMTP timeout after initial connection"

BUT....

if I send to that same external server from another account on Server#2, the email at externalserver.sbs2011.domain is processed / received ok.

Likewise if I manually utilise telnet to create an email at externalserver.sbs2011.domain it works fine.

I have been staring at this and testing as time permits for over week.

Its time to ask the audience for clues

Thanks in advance.

Tony

 

[cPRex]

Hey there! It sounds like there is a cached failure for that domain in Exim. Can you try running these commands to clear the retry database to see if that gets things working?

/usr/sbin/exim_tidydb -t 0d /var/spool/exim retry > /dev/null
/usr/sbin/exim_tidydb -t 0d /var/spool/exim reject > /dev/null
/usr/sbin/exim_tidydb -t 0d /var/spool/exim wait-remote_smtp > /dev/null

 

[thowden]

Hi cPRex

I ran the three commands, without the dev/null just to see what was processed.

The first one cleared a heap of 'too old' entries.

The last one completed with:
# /usr/sbin/exim_tidydb -t 0d /var/spool/exim wait-remote_smtp
Tidying Exim hints database /var/spool/exim/db/wait-remote_smtp
Tidying complete

The middle one seems to want something different for its syntax:
# /usr/sbin/exim_tidydb -t 0d /var/spool/exim reject
Usage: exim_tidydb [-t <time>] <spool-directory> <database-name>
<database-name> = retry | misc | wait-<transport-name> | callout | ratelimit | tls

I'll go an test if that made any difference.

Thanks
Tony

 

[thowden]

Hi

No apparent difference after running those commands on both servers.

Tony

 

[cPRex]

I'm not really sure what could be causing that then - it would be best to have our team take a look directly at the server so we can see what's going on with that. If you're able to submit a ticket to our team, please post the number here so I can follow along and make sure this thread stays updated.