Q: How to Fix Wrongly Entered External Email Forwarding Address?
-
Sep 14, 2013 5:19 PMby MrHoffman,OS X Server mail server can be configured to relay, but enabling email forwarding (autoforwarding) (HT5032) is probably a better option.
-
Sep 14, 2013 6:09 PMby icouto,Definitely do NOT want to open the mail server for relay! *shudder*
I was trying to do the autoforwarding using the GUI, and that's when I got into trouble. As far as I can tell, following the GUI instructions from the quoted knowledgebase article above, you can only autoforward to other users in the same machine - ie., to other accounts held by the same mail server.
Is that correct, or did I miss an obvious way to autoforward mail to an external address using the GUI?
In any case, I think I'm starting to see where the problem may be: reading up on the set_user_mail_opts command mentioned in the article above, I ran the following command, with the following result:
administrator$ sudo set_user_mail_opts -f, --auto_fwd list local user auto-forward settings -------------------------------- user: mary: <anne@theirserver.com>
user: postmaster: <admin@theirserver.com>
user: webmaster: <admin@theirserver.com>
user: info: <john@theirserver.com>
user: (null): <support@ourserver.com>
There seems to be a 'null' user, that is still 'forwarding' to "support@ourserver.com". Could that be the cause of the problem - and if so, how do I get rid of it?...
-
Sep 14, 2013 7:52 PMby icouto,Has (other) outbound mail from the server been tested, and reached the target server successfully?
I have just created a new test account, and tried sending mail. It seems that mail is functioning fine, and all messages reach their destination, except for messages that are sent to "ourserver.com". It seems that any messages from their server to ours - regardless of what account it's from or to - just doesn't get here.
Anything in the logs around the mail message failure?
I'm posting the most recent log dump below. The domain names and IP addresses have been changed to protect the innocent:
Sep 14 22:32:02 theirserver.com postfix/smtp[80390]: connect to mail.ourserver.com[123.45.67.89]:25: Operation timed out
Sep 14 22:32:02 theirserver.com postfix/smtp[80390]: 42F8D8012F: to=<support@ourserver.com>, relay=none, delay=30, delays=0/0/30/0, dsn=4.4.1, status=deferred (connect to mail.ourserver.com[123.45.67.89]:25: Operation timed out)
Sep 14 22:34:27 theirserver.com postfix/qmgr[80065]: 68D818009D: from=<support@ourserver.com>, size=1123, nrcpt=1 (queue active)
Sep 14 22:34:27 theirserver.com postfix/qmgr[80065]: BC7A37AE21: from=<support@ourserver.com>, size=1123, nrcpt=1 (queue active)
Sep 14 22:34:57 theirserver.com postfix/smtp[80511]: connect to mail.ourserver.com[123.45.67.89]:25: Operation timed out
Sep 14 22:34:57 theirserver.com postfix/smtp[80512]: connect to mail.ourserver.com[123.45.67.89]:25: Operation timed out
Sep 14 22:34:57 theirserver.com postfix/smtp[80511]: 68D818009D: to=<support@ourserver.com>, relay=none, delay=500, delays=470/0.02/30/0, dsn=4.4.1, status=deferred (connect to mail.ourserver.com[123.45.67.89]:25: Operation timed out)
Sep 14 22:34:57 theirserver.com postfix/smtp[80512]: BC7A37AE21: to=<support@ourserver.com>, relay=none, delay=63244, delays=63214/0.02/30/0, dsn=4.4.1, status=deferred (connect to mail.ourserver.com[123.45.67.89]:25: Operation timed out)
Sep 14 22:35:52 theirserver.com postfix/anvil[80345]: statistics: max connection rate 1/60s for (smtpd:98.76.54.32) at Sep 14 22:26:37
Sep 14 22:35:52 theirserver.com postfix/anvil[80345]: statistics: max connection count 1 for (smtpd:98.76.54.32) at Sep 14 22:26:37
Sep 14 22:35:52 theirserver.com postfix/anvil[80345]: statistics: max cache size 1 at Sep 14 22:26:37
I should also mention, that following information in the man page for the set_user_mail_opts command, I found the .mailusersettings.plist file, deleted the entry for the 'null' user, and restarted the mail server. The testing above was done after these changes had already been made...
Any suggestions would be most welcome.
-
Sep 15, 2013 9:02 AMby MrHoffman,You need to find out why the mail server is receiving those timeouts. That could be either from the mail server setup, from some IP-level routing or firewall issue, or from problems with the DNS setup or the DNS cache.
Easiest way will probably be to get somebody to ARD screen share into the box, and have a look around.
Failing that (and particularly if the environment isn't something that you'd consider "pristine"; if it's been tweaked and retweaked in some key areas), nuke and pave the box, and (re)do the minimum to get the box configured and online as the end-user of the system require it.
ps: A more general request: please don't obfuscate with real IP addresses, nor with real and registered domains. Use example.com, example.org or example.net as domains, and use completely bogus IP addresses. Thanks!
-
Sep 15, 2013 8:23 PMby icouto,MrHoffman, following your troubleshooting suggestions, I disabled the firewal on both servers, then connected remotely to "theirserver" and flushed the DNS cache - twice, just for good measure. I then checked that it could reach the data centre's DNS server, and that it could resolve "ourserver" to the correct IP address. It could.
I then tried pinging 'ourserver' from 'theirserver'. Request time out. I tried pinging other servers from "theirserver" - other clients' domains in the same datacentre, other domains in other providers, etc. Everything works flawlessly, the pings happen with no packet loss. The *only* domain it has difficulty with is "ourserver". It cannot ping it, it cannot view its webpage in Safari, it cannot access it via fileshare, ftp, or using any other protocol. It's as if the IP address for 'ourserver' has been blacklisted or blocked somewhere internally in the machine.
Is there perhaps a DNS or routing config file somewhere, that could have been messed up when I set the autoforwarding, and which could be causing this 'blockage'?
Where should I start looking?...
-
Sep 15, 2013 11:02 PMby icouto,Since writing the post above, I tried having a look at the routing tables - open Network Utility, go to the 'Netstat' tab. The first time I looked, the routing table seemed 'clean' - ie., there were no entries there that related to 'ourserver'.
A couple of minutes later, I checked the routing tables again, and saw 'ourserver' listed there, like this:
Routing tables
Internet:
Destination Gateway Flags Refs Use Netif Expire
default 888.88.88.1 UGSc 17 37 en0
127 127.0.0.1 UCS 0 0 lo0
127.0.0.1 127.0.0.1 UH 8 2125626 lo0
888.88.88 link#4 UCS 4 0 en0
888.88.88.1 0:0:c:7:ac:64 UHLWIir 18 0 en0 1189
888.88.88.2 0:8:a4:e:3b:40 UHLWIi 0 0 en0 1196
888.88.88.3 0:d0:6:f3:e0:0 UHLWIi 0 0 en0 1087
888.88.88.19 127.0.0.1 UHS 0 1487 lo0
777.77.77.68 link#4 UHRLWIi 1 9 en0 13
Where "777.77.77.68" is the IP address of 'ourserver'.
Waiting another couple of minutes, and checking the routing tables again, this entry is gone. Something seems to be adding and removing the entry to the routing table.
The most obvious part of Server that I thought could be causing this via some misconfiguration would be Mail. I made a copy of the /Library/Server/Mail/Config folder, and grepped it for 'ourserver' and its ip address - but came up empty.
What else can I try?... Any suggestions would be most welcome.
-
Jun 11, 2014 2:15 AMby petrahu,May be is to late:
/Applications/Server.app/Contents/ServerRoot/usr/bin/set_user_mail_opts -f list | grep null
user: (null): <false@mail.address.tld>
Check the contents of /Library/Server/Mail/Data/db/.mailusersettings.plist
Search the uncorrect mail addresses.
Search the correspondending user-guid, like 1AB9F9F6-024E-4D1F-9623-D22BDD93B826
<key>1AB9F9F6-024E-4D1F-9623-D22BDD93B826</key> <dict> <key>kAutoForwardValue</key> <string>false@mail.address.tld</string> <key>kMailAccountState</key> <string>Forward</string> </dict> Run:
/Applications/Server.app/Contents/ServerRoot/usr/bin/set_user_mail_opts --auto_fwd reset -g 1AB9F9F6-024E-4D1F-9623-D22BDD93B826
Regards. Petra.
