I am running ISPConfig (latest stable version) with postfix (maildir is checked). Under Admin I have created one reseller, no clients no sites. The single reseller has 3 clients and a long list of sites. Now, this is what happens: For most of the sites (but not all of them) their mails get auto-forwarded to site#1. The original receipient gets no mail at all. This is how the mail headers look like: Return-Path: <[email protected]> X-Original-To: [email protected] Delivered-To: [email protected] Received: from SendersHost (senders.host.com [senders.ip]) by client1.com (Postfix) with ESMTP id 15CDC1EECC72 for <[email protected]>; Mon, 16 Feb 2009 11:26:20 +0100 (CET) From: "sender.com" <[email protected]> To: <[email protected]> Subject: test Date: Mon, 16 Feb 2009 11:26:19 +0100 Message-ID: <001101c99021$01e894f0$05b9bed0$@com> MIME-Version: 1.0 X-Security: message sanitized on MyServer See http://www.impsec.org/email-tools/sanitizer-intro.html for details. $Revision: 1.151 $Date: 2006-01-20 07:29:24-08 X-Security: The postmaster has not enabled quarantine of poisoned messages. Content-Type: multipart/alternative; boundary="----=_NextPart_000_0012_01C99029.63ACFCF0" X-Mailer: Microsoft Office Outlook 12.0 Thread-Index: AcmQIQF/7FZ1aTN7Sjm612QjJyHcqw== X-Virus-Status: No X-Virus-Checker-Version: clamassassin 1.2.4 with clamscan / ClamAV 0.94/8995/Mon Feb 16 04:40:05 2009 In this example, mail was sent to [email protected] but was actually delivered to [email protected]. This is very weird, I cannot find any reason for this behaviour. Btw: [email protected] is defined in ISP User / User & Email. CatchAll-Email is checked, all the rest is default. /var/log/mail.log contains: Feb 16 11:26:19 MyServer postfix/smtpd[27523]: connect from senders.host.com[senders.ip] Feb 16 11:26:20 MyServer postfix/smtpd[27523]: 15CDC1EECC72: client=senders.host.com[senders.ip], sasl_method=LOGIN, sasl_username=web5_admin Feb 16 11:26:20 MyServer postfix/cleanup[27527]: 15CDC1EECC72: message-id=<001101c99021$01e894f0$05b9bed0$@com> Feb 16 11:26:20 MyServer postfix/qmgr[21656]: 15CDC1EECC72: from=<[email protected]>, size=2670, nrcpt=1 (queue active) Feb 16 11:26:20 MyServer postfix/pickup[27242]: 529101EECD0A: uid=10001 from=<web4_admin> Feb 16 11:26:20 MyServer postfix/cleanup[27527]: 529101EECD0A: message-id=<[email protected]> Feb 16 11:26:20 MyServer postfix/qmgr[21656]: 529101EECD0A: from=<[email protected]>, size=384, nrcpt=1 (queue active) Feb 16 11:26:20 MyServer postfix/local[27545]: 529101EECD0A: to=<[email protected]>, relay=local, delay=0.07, delays=0.03/0.01/0/0.02, dsn=2.0.0, status=sent (delivered to command: /usr/bi n/procmail -f-) Feb 16 11:26:20 MyServer postfix/qmgr[21656]: 529101EECD0A: removed Feb 16 11:26:22 MyServer postfix/smtpd[27523]: disconnect from senders.host.com[senders.ip] Feb 16 11:26:23 MyServer postfix/local[27528]: 15CDC1EECC72: to=<[email protected]>, orig_to=<[email protected]>, relay=local, delay=3.1, delays=0.24/0.01/0/2.8, dsn=2.0.0, status=sent (delivered to command: /usr/bin/procmail -f-) Feb 16 11:26:23 MyServer postfix/qmgr[21656]: 15CDC1EECC72: removed Feb 16 11:26:51 MyServer courierpop3login: Connection, ip=[::ffff:senders.ip] Usually the problem sits in front of the computer - but I can't find the mistake I made. Can anybody help? Regards DaFux
Take a look in your aliases file, some linux distributions put there default aliases that forward emails to root. Remove the lines that you dont need and run the newaliases command.
Sorry for my late reply - didn't see there was a message... I already checked that - aliases just contains the standards (and even them I removed for testing - no change). Address reqwriting and masquerading is also off. I have checked everything and ust can't find the problem.
No, domain and hostname are not the same. The hostname is a unique name which does not appear in my domain names...
/etc/postfix/main.cf #myorigin = /etc/mailname smtpd_banner = $myhostname ESMTP biff = no # appending .domain is the MUA's job. append_dot_mydomain = no # Uncomment the next line to generate "delayed mail" warnings #delay_warning_time = 4h # TLS parameters smtpd_tls_cert_file = /etc/postfix/ssl/smtpd.crt smtpd_tls_key_file = /etc/postfix/ssl/smtpd.key smtpd_use_tls = yes smtpd_tls_session_cache_database = btree:${queue_directory}/smtpd_scache smtp_tls_session_cache_database = btree:${queue_directory}/smtp_scache # See /usr/share/doc/postfix/TLS_README.gz in the postfix-doc package for # information on enabling SSL in the smtp client. myhostname = <domain name of site #1> alias_maps = hash:/etc/aliases alias_database = hash:/etc/aliases myorigin = /etc/mailname #mydestination = <domain name of site #1>, <hostname>, localhost.localdomain, localhost relayhost = mynetworks = 127.0.0.0/8 mailbox_command = procmail -a "$EXTENSION" mailbox_size_limit = 0 recipient_delimiter = + inet_interfaces = all inet_protocols = all smtpd_sasl_local_domain = smtpd_sasl_auth_enable = yes smtpd_sasl_security_options = noanonymous broken_sasl_auth_clients = yes smtpd_recipient_restrictions = permit_sasl_authenticated,permit_mynetworks,reject_unauth_destination smtpd_tls_auth_only = no smtp_use_tls = yes smtp_tls_note_starttls_offer = yes smtpd_tls_CAfile = /etc/postfix/ssl/cacert.pem smtpd_tls_loglevel = 1 smtpd_tls_received_header = yes smtpd_tls_session_cache_timeout = 3600s tls_random_source = dev:/dev/urandom virtual_maps = hash:/etc/postfix/virtusertable mydestination = /etc/postfix/local-host-names /etc/postfix/local-host-names this is a very long file, containing all domains on the server. Sorry, but can't publish here without causing problems with privacy and my customers in general, the file looks like localhost <hostname> localhost.<hostname> localhost.localdomain <www.site1.tld> <www.site2.tld> etc. /etc/postfix/virtusertable same problem as above - privacy, can't post that... anyway, I have checked and rechecked this file for possible problems myself many times and could not find any. All associations are correct and exactly what they should be. Thanks DaFux
Update: In my original posting I said "for most of the sites (but not all of them) their mails get auto-forwarded to site#1". I just verified that - actually all mail for all sites is forwarded to site #1.
Unfortunately not. Actually the variable $mydomain does not exist at all in my main.cf. Instead I have $myhostname which is indeed set to the domain name of the user who receives all mail. BUT: If I change that to another value and reload postfix, nothing changes. Still the same problem. The header looks like: Code: X-Original-To: info@<recipient.com> Delivered-To: web4_admin@<wrong_recipient.com> Received: from ... (... [213.xxx.xxx.xxx]) by <value in $myhostname> (Postfix) with SMTP id 75C2A1EECCE5 I tried almost everything for $myhostname, does not make any difference. Actually $myhostname is used for the server banner only, so it should not make much difference, it just appears in the mail header. Cheers DaFux
Yes, /etc/mailname contains the domain to which all mail is being forwarded. I have already replaced the line myorigin = /etc/mailname by myorigin = <domainname> (where <domainname> is the same as in myhostname) in main.cf - without any results (I reloaded/restarted postfix on the way). I just tried something else - I have set both (myorigin and myhostname to 'localhost') AND THAT SEEMED TO SOLVE THE PROBLEM. I am however worried if this setting does not have other side effects. Dor example a telnet <domainname> 25 shows this output Trying <IPAddress>... Connected to <hostname>. Escape character is '^]'. 220 localhost ESMTP QUIT 221 2.0.0 Bye Connection closed by foreign host. That "220 localhost" does not look right... However, the problem itself seems to be solved - thank you very much for your support! Cheers DaFux
OK, the original problem was solved, but now I have a new one: Mails sent by my server are coming back for several reasons. A few are: 550-HELO/EHLO must contain YOUR OWN FQDN or IP literal; localhost is MY 550-name/IP! 550 Please see RFC 2821 section 4.1.1.1 550 Access denied - Invalid HELO name (See RFC2821 4.1.1.1) (in reply to MAIL FROM command) 504 5.5.2 <localhost>: Helo command rejected: need fully-qualified hostname OK, this is the localhost which as I thought cannot be right. Question: Is it possible to make postfix use the proper domain name for the corresponding domain? I.e. when sending/receiving mail it should use the proper domain name, not just a single one (as before) or a wrong one (the localhost test - my fault).
Every mailserver uses its own hostname when sending emails and not a virtual domain. Thats the normal behaviour. Just make sure that you set the complete full hostname of the server, e.g. server1.yourdomain.com in main.cf as myhostname and mydomain and also in the file /etc/myilname in case that it exists.
That does not work ... If I use a full hostname like myserver.domainname.com, I am getting the following error: Mar 18 09:24:00 myserver postfix/smtp[16698]: 6A8931EECDAB: to=<[email protected]>, orig_to=<info@domain_belonging_to_web20>, relay=none, delay=0.04, delays=0.02/0/0.02/0, dsn=5.4.6, status=bounced (mail for myserver.domainname.com loops back to myself)
This message means that it works. You just have to set the hostname also at the end of /etc/postfix/local-host-names
Absolutely perfect! Thanks!!! That problem has taken so much time, I am really thankful that you guys did not give up on me!
