Cannot receive Mail form external Servers

[Stephan Krebernik]

It seems I cannot receive external emails anymore on my mailserver.

and haraka-smtp doesn't seem to run properly. I can't find any logs in the /log directory for haraka.

[Stephan Krebernik]

Okay. One Mailbox' Quota exceeded and because of this no other mailbox received mails anymore.

major bug.

[SH]

Do you have enough free space? Quota should not stop whole server

You can send me tail of the log from /log/s6/haraka-smtp/current to [e-mail]

[Stephan Krebernik]

Yes. Space is available. After I changed the Quota the server immediately accepted mails and delivered them to the boxes.

[SH]

I am unable to replicate that bug. Also quotas are autotested every build and there seems no problem either.

Do you have any special settings? Or do you extend container?

[Stephan Krebernik]

I had one Mailbox with 20GB Quota. This Mailbox was full. Then the strange behaviour started. After enlarging the quota to 25GB the mails started coming in again.

The Senders got the error 451 back. "Connection refused by "

I have no special settings. and I have not extended the container.

[Foddy]

I am facing the same issue since 2.0.11 PRO and I have no limits set and more than 2 TB free space on my disk... ?

This happens once in 3 days and I have to restart the container to make it work again. When I check the mail Server through a testing tool, it says timeout during this error?

/etc/cron.daily/check-user-quotas:
06:44:50 ERROR     [console] Error thrown while running command "check:user-quotas". Message: "The command "/usr/bin/doveadm -f flow quota get -A|grep -v "Limit=-"" failed.

Exit Code: 1(General error)

Working directory: /

Output:
================


Error Output:
================
" ["exception" => Symfony\Component\Process\Exception\ProcessFailedException { …},"command" => "check:user-quotas","message" => """  The command "/usr/bin/doveadm -f flow quota get -A|grep -v "Limit=-"" failed.\n  \n  Exit Code: 1(General error)\n  \n  Working directory: /\n  \n  Output:\n  ================\n  \n  \n  Error Output:\n  ================\n  """] []

In Process.php line 239:

 The command "/usr/bin/doveadm -f flow quota get -A|grep -v "Limit=-"" faile  
 d.                                                                           

 Exit Code: 1(General error)                                                  

 Working directory: /                                                         

 Output:                                                                      
 ================                                                             


 Error Output:                                                                
 ================                                                             


check:user-quotas [-h|--help] [-q|--quiet] [-v|vv|vvv|--verbose] [-V|--version] [--ansi] [--no-ansi] [-n|--no-interaction] [-e|--env ENV] [--no-debug] [--] <command>

/etc/cron.daily/logrotate:
error: rsyslog:13 duplicate log entry for /var/log/messages
error: failed to compress log /var/log/nginx/access.log.1
error: failed to compress log /var/log/rspamd/rspamd.log.1
error: failed to compress log /var/log/syslog.1
run-parts: /etc/cron.daily/logrotate exited with return code 1

[SH]

ref #366 ignore empty output return status

→ <<cset f783ec071ee1>>

[SH]

Although I am still short any clues why it is happening please try latest version, there are some minor changes.

[Foddy]

I am sorry – These kinds of errors suck... This is what I get with the last version 2.0.12.

Let me know if you need some specific logs or anything similar?

Would it be possible to implement something like the node package PM2 or node-forever, which automatically restarts the respective areas when they abort? I know this is not the solution for the problem, but maybe it gives a bit more security, especially with systems that always have to work?

[SH]

I see you are using different domain for email services and administration. Will fix that in next version...

S6 supervisor automaticaly restarts node if it crashes. Also Haraka itself handle threads in same way. I suspect this is some kind of nasty callback stuck error

[SH]

ref #366 better error handling for internal quota socket

→ <<cset 91711305240a>>

[SH]

ref #366 use different host for services vs. admin tests

→ <<cset 8125fa0243c9>>

[SH]

ref #366 use Haraka's line_socket for quota communication instead of raw nodejs net.socket

→ <<cset d80713e10b84>>

[SH]

(you can also try poste.io/mailserver:development image if you don't want to wait for release)

[Roman]

I have same issue on 2.0.11 PRO (reported by email already). I think there is some infinite loop maybe in antispam or sieve. When it happened by server has 100% CPU load and in syslog i can fidt that it's trying to deliver some message marked as *SPAM* (matched sieve filter on it) in loop.

[SH]

ref #366 failing LMTP delivery keeps crashing worker

→ <<cset da670eb32606>>

[SH]

It is most likely fixed now, please update your images

[Foddy]

Thank you! Is it the dev image or the latest 2.0.12?

[SH]

2.0.13 or dev