Concept configuration of Postfix with Smarthosts

hispeed

I have now the configuration in Fetchmail from Gerald but unfortunately i recieve this:

1 message for USERNAME at webXXX.login-12.hoststar.ch (6689 octets).
fetchmail: MDA returned nonzero status 75
reading message webXXXX@webXXX.login-12.hoststar.ch:1 of 1 (6689 octets) not flushed

My questions do I need perl for Fetchmail? Something is probably wrong in the config but which config?
I don’t get more output from Fetchmail.

martimcfly

@hispeed Hej hispeed! Maybe this is something for you?

I created a distribution for ARMv7 (RaspberryPi, Odroid) serving all your wishes. Packages for x64 are going to be build after testing.

It brings all configuration files and a customized Postfixadmin (managing Kopano and Postfix simultaneously) you can create a standalone mailserver or one which is sending with smarthost (User- or Domainwide) at the same time. Fetchmail is working, too.

The installscripts create the ready and running server. All together your done in about 15 minutes.

Introduction (Zarafa based)
https://pietma.com/zarafa-packages-for-arch-linux-arm/
https://pietma.com/zarafa-and-postfix-management/

KOPANO BUILD
https://pietma.com/run-and-access-zarafa/#comment-1973

Current image with Kopano 8.6.1
https://repository.pietma.com/nexus/content/sites/archlinux/os/prepared-kopano/

Marti

hispeed

Thanks for your post martimcfly. Actually I have a Odroid-C1 but I use it for something different. So I will need the X64 image. But I hope I can get it running on Ubuntu but if don’t get it work I will try your image as soon as it’s available for my platform based on esxi 6.5.

Gerald

@hispeed said in Concept configuration of Postfix with Smarthosts:

I have now the configuration in Fetchmail from Gerald but unfortunately i recieve this:
1 message for USERNAME at webXXX.login-12.hoststar.ch (6689 octets).
fetchmail: MDA returned nonzero status 75
reading message webXXXX@webXXX.login-12.hoststar.ch:1 of 1 (6689 octets) not flushed
My questions do I need perl for Fetchmail? Something is probably wrong in the config but which config?
I don’t get more output from Fetchmail.

If you use the way I’m using fetchmail and have suggested it above, postfix should not be involved since it will fetch the mail and directly deliver it using kopano-dagent. So it should work even if your postfix configuration is not correct (which is probably the case :-)

As fbartels has mentioned, make sure you change zarafa-dagent to kopano-dagent. My example was copy&pasted from my installation logs and it was still called zarafa at the time.

So assuming you are now using

poll pop.srv.de proto pop3 user "mylogin" pass "mypass" options ssl smtpaddress localhost forcecr mda "/usr/bin/kopano-dagent -s the-local-kopano-user-id

as your configuration, your problem is most likely with kopano-dagent since this is the MDA (mail delivery agent) and your fetchmail complains about an MDA problem. To my knowledge fetchmail does not need perl, and the error message does not in any way suggest that perl would be a problem. I suggest you first check if the executable /usr/bin/kopano-dagent specified is correct (could be in another path on other distros? I’m using ubuntu 16.04) and if this is correct you need to check the logs of kopano-dagent why it fails.

To do so, on modern systems with systemd init and current kopano configuration files you need to check the systemd log (journalctl -u kopano-dagent -f -n 20 for example). On older crappy obsolete legacy systems without systemd+journal check /var/log/kopano/ for a dagent log file. You might need to turn up the logging level if not enough can be seen in the log why it fails. This can of course be done in /etc/kopano/dagent.cfg

If you use fetchmail the way fbartels has suggested he prefers, this is of course also perfectly fine. But then your postfix server must know for which emails he is directly responsible. Your postfix log above looked like postfix sent the email on to the internet because it did not think it is responsible to locally handle it. There are many ways to configure postfix, I’m using the virtual_ directives to tell postfix for what it is responsible. But there are other ways. If you want to go this route, the options “virtual_mailbox_domains”, “virtual_mailbox_maps”, “virtual_alias_maps” and “virtual_transport” are the most impotant. You need to really learn how to configure postfix though. If you just want to fetch some mails using POP or IMAP and Postfix will alway only be used to send mail to external recipients, using fetchmail + kopano-dagent will be much easier for you.

thctlo

@martimcfly are these scripts anywhere on github? and GPL?
Then im not inventing the wheel again, this is a thing i want in debian also.
The source code page says 404 to me :-(

martimcfly

Sorry for the 404. My Gitlab (SCM) wasn’t updated for a while and I’ve made it private for security reasons. Temporaly of course.

The bash-scripts are packed in here…

/usr/share/doc/kopano/pietma
https://repository.pietma.com/nexus/service/local/repositories/archlinux/content/armv7h/pietma-kopano/kopano-core-8.6.1-1-armv7h.pkg.tar.xz

/usr/share/doc/kopano-postfixadmin/pietma/
/usr/share/doc/kopano-postfixadmin/example-config/
https://repository.pietma.com/nexus/service/local/repositories/archlinux/content/armv7h/pietma-kopano/kopano-postfixadmin-0.45.3-4-any.pkg.tar.xz

Please be aware of the fact, that these script have dependencies and asume paths which are only guaranteed with that packages on Archlinux.

But you can get a good idea how all this have to be setup to be working.

fbartels

When talking about easier user management and appliances I have to throw my “go to” solution Univention into the pot as well: https://www.univention.com/products/univention-app-center/app-catalog/kopano-core/

They also have a Fetchmail app so everything can be done from their web ui.

thctlo

@martimcfly thank you very much for these very usefull.

@fbartels, thank for the univention link, i’ll have a look at that also.

hispeed

Hi folks, thanks for your answer…slowly I’m getting closer to the solution.
The problem is the dagent. I can’t start it. Is there anything to do in the dagent.cfg? Everything was commented out (auskommentiert).

Maybe I can find out more…

Mai 30 17:19:27 svgwma-kopa-02 kopano-dagent[17333]: Wed May 30 17:19:27 2018: [info   ] Coredump status left at system default.
Mai 30 17:19:27 svgwma-kopa-02 kopano-dagent[17333]: Wed May 30 17:19:27 2018: [info   ] Maximum LMTP threads set to 20
Mai 30 17:19:27 svgwma-kopa-02 kopano-dagent[17333]: Wed May 30 17:19:27 2018: [info   ] Listening on *:2003 for LMTP
Mai 30 17:19:27 svgwma-kopa-02 kopano-dagent[17333]: Wed May 30 17:19:27 2018: [crit   ] Unable to create socket(10,1,6) port 2003: Address already in use

https://forums.zarafa.com/showthread.php?12412-zarafa-7-2-2-dagent-error

Same here is there any sample for dagent.cfg?
@thctlo i’ts always comming back ;=) 2 years are over and nobody has the problem except from me and you had it? How did you fix it?

/etc/kopano/server.cfg <- I have probably some configuration issues also there.

For me it looks like that this tutorial is incomplete as always:
https://www.pc-howto.com/kopano-der-neue-stern-am-groupwarehimmel-teil-1/
https://www.pc-howto.com/kopano-der-neue-stern-am-groupwarehimmel-teil-2/

Maybe I’m just to stupid. @martimcfly I go thru your config files by tomorrow.

Gerald

@hispeed

If you run kopano-dagent with the parameter -l, it will bind to port 2003 and wait for connections. [-l = listen mode]

If you start kopano-dagent with a receipient as parameter it will read mail from stdin and then (with proper input which it gets from fetchmail) deliver this single mail to kopano.

If your log says “Unable to create socket, address already in use”, then you already have a kopano-dagent running in listen mode and you are trying to start another one in listen mode. kopano-dagent cannot run twice in listen mode. But you can of course have it running in listen mode and on demand call it in interactive mode from fetchmail.

An empty dagent.cfg should be fine.

hispeed

Hi Gerald,

Now I’m more confident with logs and dagent. It looks now fine.

I think that I have a problem in fetchmailrc:

#.fetchmailrc
set syslog;
set postmaster "email_from_root_user_i_created_in_kopano";
# set no bouncemail
set no softbounce;
# accounts
poll webXXX.login-XXX.hoststar.ch proto pop3 user webXXXX pass XXXXXXX options ssl smtpaddress localhost forcecr mda usr/sbin/kopano-dagent -s 4

This is wrong because of “-s” after kopano-dagent. Do I have to use " " in the syntax if yes, only at mda “usr/sbin/kopano-dagent -s 4”?
Number 4 is because it’s the sql table written number 4 can I use the kopano-login name as well?

When I use it with the username and quotes " " and I have to use the directory “usr/sbin/kopano-dagent -s username”
in “bin” I don’t have a kopano-dagent only in “sbin”.
I always recieve:

root@svgwma-kopa-02:/home/kopano# tail -f /var/log/mail.log
May 31 07:23:46 svgwma-kopa-02 fetchmail[3703]: 1 message for webXXXX at webXXX.login-XX.hoststar.ch (6722 octets).
May 31 07:23:47 svgwma-kopa-02 fetchmail[3703]: reading message webXXXX@webXXX.login-XX.hoststar.ch:1 of 1 (6722 octets) (log message incomplete) <--- WHY Incomplete??? 
May 31 07:23:47 svgwma-kopa-02 fetchmail[3703]: MDA returned nonzero status 75

I use the newest Ubuntu server version…

On the other side Kopano Core Delivery Agent with Debug Mode:
The only thing I see which is not correct, is the time. Actually I don’t know why.

EDIT 3:

Mai 31 07:28:03 svgwma-kopa-02 systemd[1]: Started Kopano Core Delivery Agent.
Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [info   ] [ 3816] Logger process started on pid 3845
Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [debug  ] [ 3816] StatsClient binding socket
Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [debug  ] [ 3816] StatsClient bound socket to /tmp/.5bc7b9886cc74b9a.sock
Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [debug  ] [ 3816] StatsClient thread started
Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [=======] [ 3816] Starting kopano-dagent version 8.6.80 (pid 3816) (LMTP mode)
Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [debug  ] [ 3816] Submit thread started

When I want to list all Kopano users I get this:

root@svgwma-kopa-02:/home/kopano# kopano-cli --list-users
Traceback (most recent call last):
  File "/usr/sbin/kopano-cli", line 3, in <module>
    import kopano_cli
  File "/usr/lib/python2.7/dist-packages/kopano_cli/__init__.py", line 10, in <module>
    from MAPI.Tags import PR_EC_STATSTABLE_SYSTEM, PR_DISPLAY_NAME, PR_EC_STATS_SYSTEM_VALUE
ImportError: No module named MAPI.Tags
root@svgwma-kopa-02:/home/kopano# kopano-cli
Traceback (most recent call last):
  File "/usr/sbin/kopano-cli", line 3, in <module>
    import kopano_cli
  File "/usr/lib/python2.7/dist-packages/kopano_cli/__init__.py", line 10, in <module>
    from MAPI.Tags import PR_EC_STATSTABLE_SYSTEM, PR_DISPLAY_NAME, PR_EC_STATS_SYSTEM_VALUE
ImportError: No module named MAPI.Tags
root@svgwma-kopa-02:/home/kopano# kopano-cli

Something is here wrong with the DB or so…

Gerald

@hispeed said in Concept configuration of Postfix with Smarthosts:

I think that I have a problem in fetchmailrc:

This is wrong because of “-s” after kopano-dagent. Do I have to use " " in the syntax if yes, only at mda “usr/sbin/kopano-dagent -s 4”?
Number 4 is because it’s the sql table written number 4 can I use the kopano-login name as well?

Ok, I had “the-local-kopano-user-id” in my example. But that did not mean that you need to dig in the MySQL Database and search for a random hidden usernumber. What is meant by this is just the Username in Kopano.
I don’t know about the new kopano-cli but in kopano-admin this is what is specified as -u and what is listed as Username in kopano-admin -l

When I use it with the username and quotes " " and I have to use the directory “usr/sbin/kopano-dagent -s username”

Why no / in front of /usr? Might work depending on where it is run from, but this looks weird and is an unnecessary reason for it to fail.

in “bin” I don’t have a kopano-dagent only in “sbin”.
I use the newest Ubuntu server version…

Yeah, just mentioned it because this could have been different in other distros. But I’m also using Ubuntu and sbin is correct. Easy to check:

root@myhostname:~# whereis kopano-dagent
kopano-dagent: /usr/sbin/kopano-dagent /usr/share/kopano-dagent /usr/share/man/man8/kopano-dagent.8.gz

I always recieve:

May 31 07:23:47 svgwma-kopa-02 fetchmail[3703]: reading message webXXXX@webXXX.login-XX.hoststar.ch:1 of 1 (6722 octets) (log message incomplete) <--- WHY Incomplete??? 
May 31 07:23:47 svgwma-kopa-02 fetchmail[3703]: MDA returned nonzero status 75

I have no idea. Perhaps remove the -s from the kopano-dagent parameter (this is for silent)
Instead add -v (verbose) or -vv (even more verbose)!

Also add verbose flags to fetchmail, Syntax probably the same (-v or -vv for even more verbosity).

On the other side Kopano Core Delivery Agent with Debug Mode:
The only thing I see which is not correct, is the time. Actually I don’t know why.

I think I had this once too in some log… Don’t remember exactly, but check timedatectl if your timezones are correct.

Output for me is:

root@somehostname:~# timedatectl
      Local time: Thu 2018-05-31 23:28:08 CEST
  Universal time: Thu 2018-05-31 21:28:08 UTC
        RTC time: Thu 2018-05-31 21:28:08
       Time zone: Europe/Berlin (CEST, +0200)
 Network time on: yes
NTP synchronized: yes
 RTC in local TZ: no

Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [=======] [ 3816] Starting kopano-dagent version 8.6.80 (pid 3816) (LMTP mode)
Mai 31 07:28:03 svgwma-kopa-02 kopano-dagent[3816]: Thu May 31 07:28:03 2018: [debug  ] [ 3816] Submit thread started

This looks like the log from the zarafa-dagent in daemon mode. (Parameter -s)
Not sure if the logs from a manually started zarafa-dagent in non-daemon mode will end up here to. At least nothin in the log you quoted.

When I want to list all Kopano users I get this:

root@svgwma-kopa-02:/home/kopano# kopano-cli --list-users
Traceback (most recent call last):
  File "/usr/sbin/kopano-cli", line 3, in <module>
ImportError: No module named MAPI.Tags
root@svgwma-kopa-02:/home/kopano# kopano-cli

Something is here wrong with the DB or so…

Does kopano-admin -l work for you? The error message looks more like you have not installed a required package?

hispeed

Hi,

With verbose I recieve this:

fetchmail@svgwma-kopa-02:~$ fetchmail -f fetchmail-accounts
Sat Jun  2 07:43:32 2018: [info   ] Coredump status left at system default.
Sat Jun  2 07:43:32 2018: [11417] [debug  ] StatsClient binding socket
Sat Jun  2 07:43:32 2018: [11417] [debug  ] StatsClient bound socket to /tmp/.32877d08a35722c.sock
Sat Jun  2 07:43:32 2018: [11417] [debug  ] Submit thread started
Sat Jun  2 07:43:32 2018: [11417] [debug  ] StatsClient thread started
Sat Jun  2 07:43:32 2018: [11417] [debug  ] PYTHONPATH = /usr/share/kopano-dagent/python
Sat Jun  2 07:43:32 2018: [11417] [error  ]   Python type: (null)
Sat Jun  2 07:43:32 2018: [11417] [error  ]   Python error: No module named MAPI
Sat Jun  2 07:43:32 2018: [11417] [crit   ] K-1732: Unable to initialize the dagent plugin manager: Unknown error code (1).
Sat Jun  2 07:43:32 2018: [11417] [debug  ] StatsClient terminating
Sat Jun  2 07:43:32 2018: [11417] [debug  ] StatsClient terminated

I will check if I can update Kopano and test again…

EDIT 1:

I download via wget the newest kopano-core and then made the package and installed it.
This worked fine so far. After the installation the system user on Ubuntu 18.04 had a new Password <- why?

Now when I want to login via Webapp I recieve this error:

Unknown MAPI Error: MAPI_E_NOT_FOUND

When I make this: kopano-admin --create-store hispeed

Then I recieve:

kopano-admin: relocation error: kopano-admin: symbol _ZN2KC21GetAutoAcceptSettingsEP9IMsgStorePbS2_S2_S2_ version KC_8.6.80 not defined in file libmapi.so.1 with link time reference

It’s getting worser from day to day ;=)…

hispeed

Yes yes yes…

I don’t believe it I can now recieve. I updated the whole Ubuntu and restartet. Then I had to change something in the config from Kopano. Attachament stored as File (I made there a change earlier). I’m sure I can switch that to database later.
Restarting everything and now I can recieve e-mails.

Thanks so far for everyone. There are still a few questions open like: why the kopano user gets a new password after update?

Now I have to figure out how I can send E-mails… For this I’m going to use postfix.

Postfix i’m back in trouble :D! Yes something is wrong in my postfix configuration.

Jun  3 16:20:58 svgwma-kopa-02 kopano-server[2270]: message repeated 9 times: [ Error while connecting to search on "file:///var/run/kopano/search.sock"]
Jun  3 16:21:01 svgwma-kopa-02 postfix/pickup[10947]: A6A46320DFE: uid=0 from=<root>
Jun  3 16:21:01 svgwma-kopa-02 postfix/cleanup[10957]: A6A46320DFE: message-id=<20180603162101.A6A46320DFE@svgwma-kopa-02.mydomain.me>
Jun  3 16:21:01 svgwma-kopa-02 postfix/qmgr[10948]: A6A46320DFE: from=<root@mydomain.me>, size=727, nrcpt=1 (queue active)
Jun  3 16:21:01 svgwma-kopa-02 postfix/lmtp[10959]: host localhost[::1] offers SMTPUTF8 support, but not 8BITMIME
Jun  3 16:21:01 svgwma-kopa-02 postfix/lmtp[10959]: A6A46320DFE: to=<root@mydomain.me>, orig_to=<root>, relay=localhost[::1]:2003, delay=0.18, delays=0.03/0.01/0.06/0.07, dsn=5.1.1, status=bounced (host localhost[::1] said: 503 5.1.1 User does not exist (in reply to RCPT TO command))
Jun  3 16:21:01 svgwma-kopa-02 postfix/cleanup[10957]: CE28C320DFF: message-id=<20180603162101.CE28C320DFF@svgwma-kopa-02.mydomain.me>
Jun  3 16:21:01 svgwma-kopa-02 postfix/bounce[10962]: A6A46320DFE: sender non-delivery notification: CE28C320DFF
Jun  3 16:21:01 svgwma-kopa-02 postfix/qmgr[10948]: CE28C320DFF: from=<>, size=2791, nrcpt=1 (queue active)
Jun  3 16:21:01 svgwma-kopa-02 postfix/qmgr[10948]: A6A46320DFE: removed
Jun  3 16:21:01 svgwma-kopa-02 postfix/lmtp[10959]: host localhost[::1] offers SMTPUTF8 support, but not 8BITMIME
Jun  3 16:21:01 svgwma-kopa-02 postfix/lmtp[10959]: CE28C320DFF: to=<root@mydomain.me>, relay=localhost[::1]:2003, delay=0.11, delays=0.01/0/0.06/0.05, dsn=5.1.1, status=bounced (host localhost[::1] said: 503 5.1.1 User does not exist (in reply to RCPT TO command))
Jun  3 16:21:01 svgwma-kopa-02 postfix/qmgr[10948]: CE28C320DFF: removed

EDIT: 3
Gerald or someone else:

Do you have a sample of the postfix configuration?

martimcfly

@hispeed let me know when you need advice

thctlo

Postfix offers SMTPUTF8 support, kopano LMTP not.
Set in postfix smtputf8_enable = no restart postfix and its fixed. ;-)

hispeed

Hi martimcfly,

Yes I need help. My configuration looks like that at the moment:

# See /usr/share/postfix/main.cf.dist for a commented, more complete version

compatibility_level = 2

mydomain = mydomain.me
myorigin = $mydomain
mynetworks = 127.0.0.0/8, [::1]/128
smtp_host_lookup = dns, native

mailbox_size_limit = 0
message_size_limit = 52428800

delay_warning_time = 4h
unknown_local_recipient_reject_code = 450
maximal_queue_lifetime = 7d
minimal_backoff_time = 1000s
maximal_backoff_time = 8000s

# header_checks = regexp:/etc/kopano/postfix/header_checks
# body_checks = regexp:/etc/kopano/postfix/body_checks

alias_maps = hash:/etc/aliases
virtual_alias_maps = hash:/etc/kopano/postfix/valiases

virtual_mailbox_domains = /etc/kopano/postfix/vdomains
virtual_transport = lmtp:localhost:2003

smtpd_banner = $myorigin ESMTP
smtpd_helo_required = yes
smtpd_client_restrictions = permit_mynetworks
smtpd_recipient_restrictions = permit_mynetworks reject_invalid_hostname reject_unauth_destination reject_unknown_recipient_domain
smtpd_sender_restrictions = reject_unknown_address
smtpd_recipient_limit = 16
smtpd_soft_error_limit = 3
smtpd_hard_error_limit = 12

smtp_tls_security_level = may
smtp_sasl_auth_enable = yes
smtp_sender_dependent_authentication = yes
smtp_sasl_password_maps = hash:/etc/kopano/postfix/sasl_passwd
sender_dependent_relayhost_maps = hash:/etc/kopano/postfix/sender_relay
smtp_use_tls = yes
smtp_tls_enforce_peername = no
smtp_sasl_security_options = noanonymous
smtp_helo_timeout = 120s

# Korrekturen
smtputf8_autodetect_classes = verify

I need to use this two lines or something similar:

smtp_sasl_password_maps = hash:/etc/kopano/postfix/sasl_passwd
sender_dependent_relayhost_maps = hash:/etc/kopano/postfix/sender_relay

With this config it doesn’t work. Postfix keeps telling me that the user doesn’t exist in Kopano. But I have created a user with this e-mail.

Thanks for your ideas…

thctlo

Have you even tried what i did say… :-/

Your logs show:
host localhost[::1] offers SMTPUTF8 support, but not 8BITMIME

kopano-lmtp does NOT support SMTPUTF8 and you mail is rejected.

What you see in your logs is correct.

https://forum.kopano.io/topic/1262/kopano-smtputf8-support
Already reported this some time ago.

thctlo

and about you postfix config. this part needs serious fixing, this is not ok.

smtpd_helo_required = yes
smtpd_client_restrictions = permit_mynetworks
smtpd_recipient_restrictions = permit_mynetworks reject_invalid_hostname reject_unauth_destination reject_unknown_recipient_domain
smtpd_sender_restrictions = reject_unknown_address

I suggest and this is a verified config by the postfix list.
i have more but this is a working setup

smtpd_delay_reject = yes

# Obey the RFC's. any hostname should have an A and/or MX and/or PTR ( or resolvable CNAME ) 
# see RFC https://tools.ietf.org/html/rfc2821#section-2.3.4  and 2.3.5 
smtpd_client_restrictions =
    permit_mynetworks,
    reject_unauth_destination,
    reject_non_fqdn_hostname,
    reject_unknown_hostname,
    reject_invalid_hostname,
    reject_unauth_pipelining

smtpd_helo_required = yes
smtpd_helo_restrictions =
    permit_mynetworks,
    reject_unauth_destination,
    reject_invalid_helo_hostname,
    reject_non_fqdn_helo_hostname,
    reject_unknown_helo_hostname,
    reject_unauth_pipelining

smtpd_sender_restrictions =
    permit_mynetworks,
    reject_unauth_destination,
    reject_non_fqdn_sender,
    reject_unknown_sender_domain,
    reject_unknown_address,
    reject_unauth_pipelining

smtpd_recipient_restrictions =
    permit_mynetworks,
    reject_unauth_destination,
    reject_non_fqdn_recipient,
    check_recipient_access hash:/etc/postfix/personal/check_recipient_access-allow.map
    reject_unknown_recipient_domain,
    reject_multi_recipient_bounce,
    reject_unlisted_recipient,
    reject_unverified_recipient

smtpd_relay_restrictions =
    permit_mynetworks,
    reject_unauth_destination,
    check_recipient_access hash:/etc/postfix/personal/check_recipient_access-allow.map
    reject_multi_recipient_bounce,
    reject_non_fqdn_hostname,
    reject_invalid_hostname,
    reject_invalid_helo_hostname,
    reject_non_fqdn_sender,
    reject_non_fqdn_recipient,
    reject_unknown_sender_domain,
    reject_unknown_recipient_domain,
    defer_unauth_destination
##
smtpd_data_restrictions =
    reject_unauth_pipelining,
    reject_multi_recipient_bounce
##
smtpd_etrn_restrictions =
    permit_mynetworks,
    reject

Now, your internet ready.

hispeed

@thctlo yes this does fix this error if I add this line. But still there are some other issues.

@martimcfly I have copied now your configuration. How do I now create the check_recipient_access-allow.map?
Where do I define the passwords and login data for my different mail accounts?

At the moment I get this error:

Jun  4 17:52:35 svgwma-kopa-02 kopano-server[12291]: Error while connecting to search on "file:///var/run/kopano/search.sock"
Jun  4 17:53:25 svgwma-kopa-02 kopano-server[12291]: message repeated 5 times: [ Error while connecting to search on "file:///var/run/kopano/search.sock"]
Jun  4 17:54:01 svgwma-kopa-02 postfix/pickup[26277]: D3277320E04: uid=0 from=<root>
Jun  4 17:54:01 svgwma-kopa-02 postfix/cleanup[26505]: D3277320E04: message-id=<20180604175401.D3277320E04@svgwma-kopa-02.localdomain>
Jun  4 17:54:01 svgwma-kopa-02 postfix/qmgr[26278]: D3277320E04: from=<root@svgwma-kopa-02.localdomain>, size=757, nrcpt=1 (queue active)
Jun  4 17:54:01 svgwma-kopa-02 postfix/local[26507]: warning: dict_nis_init: NIS domain name not set - NIS lookups disabled
Jun  4 17:54:01 svgwma-kopa-02 postfix/local[26507]: D3277320E04: to=<root@svgwma-kopa-02.localdomain>, orig_to=<root>, relay=local, delay=0.06, delays=0.02/0/0/0.03, dsn=2.0.0, status=sent (delivered to mailbox)
Jun  4 17:54:01 svgwma-kopa-02 postfix/qmgr[26278]: D3277320E04: removed

EDIT:
@fbartels Can you help me with this: Jun 4 17:52:35 svgwma-kopa-02 kopano-server[12291]: Error while connecting to search on “file:///var/run/kopano/search.sock” ?
-> This is solved with: Search Socket Error

On the images you see my old Zarafa config, which is running right now. This works fine for me. Maybe there’s a security risk somewhere I don’t know. First I need to create that on Kopano and then I can add security.

0_1528135898230_2018-06-04 20_07_56-Titanserver_zarafa_config_postfix.jpg
1_1528135898231_2018-06-04 20_07_56-Titanserver_zarafa_config_postfix_2.jpg

Update:
SSL self-signed = Is working
Z-Push 2.4.2 = Is working with Ubuntu 18.04 (No official build I took: 16.04 version)
Deskapp = Is working
Cron Job for Fetchmail = Is working

To do:
Send E-Mails ;=)