Mail for root not received after latest updates

mail
v7

(Eddie Atherton) #1

NethServer Version: V7
Module: mail

Following the updates a few days ago, I am no longer receiving mail from one of my VMs addressed to root. Here’s the maillog for one of the attempts:

Jul  3 03:20:40 Nethserver postfix/smtpd[21511]: connect from FortKnox.BogoLinux.net[192.168.0.100]
Jul  3 03:20:40 Nethserver postfix/smtpd[21511]: NOQUEUE: client=FortKnox.BogoLinux.net[192.168.0.100]
Jul  3 03:20:40 Nethserver postfix/smtpd[21511]: warning: connect to proxy filter 127.0.0.1:10024: Connection refused
Jul  3 03:20:40 Nethserver postfix/smtpd[21511]: proxy-reject: END-OF-MESSAGE: 451 4.3.0 Error: queue file write error; from=<root@fortknox.bogolinux.net> to=<root@nethserver.bogolinux.net> proto=SMTP helo=<fortknox.bogolinux.net>
Jul  3 03:20:40 Nethserver postfix/smtpd[21511]: disconnect from FortKnox.BogoLinux.net[192.168.0.100]

The listener for port 10024 is a little different than the other tcp6 ones:

[root@Nethserver ~]# netstat -an | grep LISTEN\  | grep tcp6
tcp6       0      0 ::1:10024               :::*                    LISTEN
tcp6       0      0 :::587                  :::*                    LISTEN
tcp6       0      0 :::110                  :::*                    LISTEN

Note the extra “1”.

Is this the reason and how can I fix it.

** UPDATE **
Possibly not the reason, as well as amavisd (10024 above), spamd (783) and unbound (8953) also use the same addressing.

Cheers.


(Michael Träumner) #2

Please have a look at https://www.howtoforge.com/community/threads/127-0-0-1-127-0-0-1-10024-connection-refused.39453/
Perhaps one of this tips could help you.


(Eddie Atherton) #3

Previous to the last updates I was receiving mail correctly. I do not, and have not, modified any of the configuration files involved in the mail delivery.

Pointing me at a thread that implies amavisd is not running does not help, as my post clearly shows that port 10024 is in listening mode.

Cheers.


(Michael Träumner) #4

Perhaps the update did.[quote=“EddieA, post:3, topic:7295”]
Pointing me at a thread that implies amavisd is not running does not help
[/quote]

Did you read this?

Have the same issue:
Dec 19 15:23:56 sd postfix/smtp[179]: BFE3738054E: to=the@mail.com, relay=none, delay=0.14, delays=0.13/0.01/0/0, dsn=4.4.1, status=deferred (connect to 127.0.0.1[127.0.0.1]:10024: Connection refused)


>postfix and amavisd-new are started.
>Restart amavisd and if it does not start, check the mail.log file for the reason.
>I did this earlier: In /etc/clamav/clamd.conf AllowSupplementaryGroups false -> true
But then I didn't restart it... So: service clamav-daemon restart Solved the issue.

PS: I only tried to help, it’s not my area of expertise.


(Filippo Carletti) #5

Please confirm that amavisd is working on port 10024 with the following command:

$ nc localhost 10024
220 [127.0.0.1] ESMTP amavisd-new service ready

If it doesn’t, try to restart the service (through server-manager or systemctl restart amavisd
If the restart fixes, please search /var/log/maillog for amavisd restarts in the past (maybe close in time to the update).
Let us know, please.


(Eddie Atherton) #6

I’ll keep an eye on this:

[root@Nethserver ~]# netstat -an | grep 10024
tcp6       0      0 ::1:10024               :::*                    LISTEN
[root@Nethserver ~]# nc localhost 10024
Ncat: Connection refused.
[root@Nethserver ~]# systemctl restart amavisd
[root@Nethserver ~]# netstat -an | grep 10024
tcp        0      0 127.0.0.1:10024         0.0.0.0:*               LISTEN
tcp6       0      0 ::1:10024               :::*                    LISTEN
[root@Nethserver ~]# nc localhost 10024
220 [127.0.0.1] ESMTP amavisd-new service ready

The updates were applied on Jul 2 @ 12:27:

/var/log/messages:Jul  2 12:28:45 Nethserver esmith::event[15080]: expanding /etc/amavisd/amavisd.conf
/var/log/messages:Jul  2 12:28:51 Nethserver esmith::event[15080]: [INFO] service amavisd restart
/var/log/messages:Jul  2 12:28:54 Nethserver esmith::event[15080]: [INFO] amavisd restart
/var/log/messages:Jul  2 12:33:17 Nethserver systemd: Started clamd scanner (amavisd) daemon.
/var/log/messages:Jul  2 12:33:17 Nethserver systemd: Starting clamd scanner (amavisd) daemon...
/var/log/messages:Jul  2 12:37:53 Nethserver systemd: Starting Clean amavisd tmp folder...
/var/log/messages:Jul  2 12:37:53 Nethserver systemd: Starting Clean amavisd quarantine folder...
/var/log/messages:Jul  2 12:37:54 Nethserver systemd: Started Clean amavisd tmp folder.
/var/log/messages:Jul  2 12:37:54 Nethserver systemd: Started Clean amavisd quarantine folder.
/var/log/messages:Jul  3 12:37:55 Nethserver systemd: Starting Clean amavisd tmp folder...
/var/log/messages:Jul  3 12:37:55 Nethserver systemd: Starting Clean amavisd quarantine folder...
/var/log/messages:Jul  3 12:37:55 Nethserver systemd: Started Clean amavisd tmp folder.
/var/log/messages:Jul  3 12:37:55 Nethserver systemd: Started Clean amavisd quarantine folder.
/var/log/messages:Jul  4 12:37:56 Nethserver systemd: Starting Clean amavisd tmp folder...
/var/log/messages:Jul  4 12:37:56 Nethserver systemd: Starting Clean amavisd quarantine folder...
/var/log/messages:Jul  4 12:37:56 Nethserver systemd: Started Clean amavisd tmp folder.
/var/log/messages:Jul  4 12:37:56 Nethserver systemd: Started Clean amavisd quarantine folder.

I only get a couple of messages each day, both early morning and the last received were July 2. I’ll check again tomorrow.

Cheers.


(Eddie Atherton) #7

OK, so after today’s updates and re-boot, I now have this:

[root@Nethserver ~]# netstat -an | grep 10024
tcp        0      0 127.0.0.1:10024         0.0.0.0:*               LISTEN
[root@Nethserver ~]#

So, what it looks like happened was that after the re-boot for the previous updates, amavisd started on the IPv6 listener instead of the IPv4. The restart I did mid re-boots, based on the earlier replies, correctly, started it on the IPv4 which made everything work fine.

Wonder why the previous re-boot picked the wrong protocol.

Cheers.


(Michael Träumner) #8

If you have solved it, please mark your answer as solution.


(Stefano Zamboni) #9

IMO his issue is resolved, but the reason why he had such an issue must be investigated


(Filippo Carletti) #10

amavisd has been recently updated to run better under systemd (which is our/centos7 case).
I don’t know if it may have solved this issue.
I see
After=network.target
in amavisd systemd startup script.