Split DNS and Disaster Recovery and other weird issues

[skrid216]

I am in the process of trying to recover from a hardware failure (CPU/mobo).

I have successfully installed a "dummy" copy of Zimbra and restored all of the mail files.

However, I am struggling with getting mail to flow properly. I did not change the host name or configuration at all.

Here is my situation:

I originally set up Zimbra with a "dummy" domain: zimbra.datacenter.inc. This was an internal DNS scheme, meaning no one would send mail to the datacenter.inc domain. I then created users with valid domains (i.e. coopermackenzie.com, streetlevelfm.com, thedirksfamily.us etc.). Prior to the hardware failure, Zimbra was working perfectly and routing mail to these alternate domains. On the old hard drive, I can find no evidence that I had an internal DNS server set up.

However after I rebuilt a new machine, I decided to try and make it closer to what it should be. Therefore, I have been trying to follow all of the articles and advice for using BIND and Split DNS. However, I cannot get mail flowing. As long as I check the "Enable DNS lookups" checkbox, mail will be delivered to external domains. However, in order to receive inbound mail, I have to uncheck this setting and then outbound mail sits in the "Deferred" queue.

I am at my wits end and cannot figure this out. I have searched the forums and tried to follow all of the wonderful advice...to no avail.

Please help.

[phoenix]

You restore your server to the state it was in before the hardware failure and not change anything - modifications should be made after your server is working correctly.

There are posts all over the forum with the commands to diagnose the problem so here they are again. Post the output of the following commands (run on the Zimbra server):

cat /etc/hosts
cat /etc/resolv.conf
dig yourdomain.com mx
dig yourdomain.com any
host `hostname` <- use that exact command with backticks not single quotes

[skrid216]

Thanks for the help. I restored it to previous state and following many of the guides and posts for doing this. Only after doing this did I start to make changes to try to get mail flowing.

Another interesting issue is that my HTTP server is listening on port 81, not 80. I haven't dug into that one much...been focused on getting mail flowing properly.

Here is the output of the commands.

[root@zimbra ~]# cat /etc/hosts
# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1 localhost.localdomain localhost
192.168.1.10 zimbra.datacenter.inc zimbra


[root@zimbra ~]# cat /etc/resolv.conf
search datacenter.inc
nameserver 192.168.1.10
nameserver 192.168.1.254
nameserver 8.8.8.8


[root@zimbra ~]# dig datacenter.inc mx

; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> datacenter.inc mx
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 12233
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;datacenter.inc. IN MX

;; Query time: 850 msec
;; SERVER: 192.168.1.254#53(192.168.1.254)
;; WHEN: Fri Jun 11 08:23:34 2010
;; MSG SIZE rcvd: 32



[root@zimbra ~]# dig datacenter.inc mx

; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> datacenter.inc mx
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 12233
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;datacenter.inc. IN MX

;; Query time: 850 msec
;; SERVER: 192.168.1.254#53(192.168.1.254)
;; WHEN: Fri Jun 11 08:23:34 2010
;; MSG SIZE rcvd: 32

[root@zimbra ~]# dig datacenter.inc any

; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> datacenter.inc any
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 64720
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;datacenter.inc. IN ANY

;; Query time: 906 msec
;; SERVER: 192.168.1.254#53(192.168.1.254)
;; WHEN: Fri Jun 11 08:24:35 2010
;; MSG SIZE rcvd: 32


[root@zimbra ~]# host `hostname`
zimbra.datacenter.inc has address 192.168.1.10
zimbra.datacenter.inc mail is handled by 10 zimbra.datacenter.inc.

[skrid216]

So, here is some additional information.

I know this is DNS related...here are some additional symptoms:

If I uncheck the "Enable DNS lookups" in the Zimbra Admin Console, inbound mail works fine.

If I check the "Enable DNS lookups" in the Zimbra Admin Console, outbound mail works fine.

Does that make any sense?

Also, obviously the host "zimbra.datacenter.inc" does not resolve externally. That was only used as an internal host name.

The other domains for which I am hosting should resolve externally:

thedirksfamily.us
streetlevelfm.com
etc...

[skrid216]

You restore your server to the state it was in before the hardware failure and not change anything - modifications should be made after your server is working correctly.

Yes, that's what I did. I restored the server to the same version, and all of the mail restored fine. There were a few LDAP issues, but I was able to resolve those as well.

Here is some additional information.

I know this is DNS related...here are some additional symptoms:

If I uncheck the "Enable DNS lookups" in the Zimbra Admin Console, inbound mail works fine.

If I check the "Enable DNS lookups" in the Zimbra Admin Console, outbound mail works fine.

Does that make any sense?

Also, obviously the host "zimbra.datacenter.inc" does not resolve externally. That was only used as an internal host name.

The other domains for which I am hosting should resolve externally:

thedirksfamily.us
streetlevelfm.com
etc...

Post the output of the following commands (run on the Zimbra server):

[root@zimbra ~]# cat /etc/hosts
# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1               localhost.localdomain localhost
192.168.1.10            zimbra.datacenter.inc   zimbra
[root@zimbra ~]# cat /etc/resolv.conf
search datacenter.inc
nameserver 192.168.1.10
nameserver 192.168.1.254
nameserver 8.8.8.8
[root@zimbra ~]# dig datacenter.inc mx

; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> datacenter.inc mx
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 37456
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;datacenter.inc.                        IN      MX

;; Query time: 1040 msec
;; SERVER: 192.168.1.254#53(192.168.1.254)
;; WHEN: Fri Jun 11 20:57:54 2010
;; MSG SIZE  rcvd: 32

[root@zimbra ~]# dig datacenter.inc any

; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> datacenter.inc any
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 18191
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;datacenter.inc.                        IN      ANY

;; Query time: 891 msec
;; SERVER: 192.168.1.254#53(192.168.1.254)
;; WHEN: Fri Jun 11 20:58:21 2010
;; MSG SIZE  rcvd: 32

[root@zimbra ~]# host `hostname`
zimbra.datacenter.inc has address 192.168.1.10
zimbra.datacenter.inc mail is handled by 10 zimbra.datacenter.inc.

[rlomba]

Maybe your split dns is not working correctly (I see that when you "dig", you get the answer from 192.168.1.254 and I understand that's not your machine).

Edit your /etc/resolv.conf and leave just your local server as DNS.

Make sure you have the necessary stuff to make the split DNS work:
yum install bind
yum install bind-chroot

Follow the directions here to set up a correct split DNS:
Split DNS - Zimbra :: Wiki

Make sure you configure your local split-dns so as the FORWARDER is your ISP's DNS (or Public DNS, or 192.168.1.254, or whatever it should be).

Start everything:
/etc/init.d/named start
chkconfig named on

Check that it's working

HTH