Migration problem from FC3 to Centos 4.3

[kibo]

Hello All

Well! The scenario is that I have a Zimbra 3.1 installed on FC3 which is working fine. I decided to move the server to somewhere else. The new server is a Xeon with Centos 4.3 Xen Dom0 installed and another Centos 4.3 installed as guest. The Glibc libraries have been compiled and installed successfully on the guest and TLS disabled on the dom0. Apparently Zimbra needs them. Then I did following:

1- Downloaded 3.1 RHEL4 version of Zimbra and installed successfully under guest xen Centos.
2- Backed up /opt/zimbra from FC3 via tar and transferred to the same folder in the guest xen OS in the same location. (I deleted the /opt/zimbra from step1)
3- Did a zmfixperms
4- I reinstalled Zimbra 3.1 RHEL4 and asked to upgrade.
5- It successfully upgraded the Zimbra from FC3 to RHEL4 and everything looked fine to me afterwards.
6- I redirected the MX records to the new Zimbra.


Then I noticed that whatever mail I send from any account or whatever e-mail coming from outside goes directly to the deffered queue and they don't go away by flushing. In the error section of the admin panel, it says that the delivery is temporary suspended because host 127.0.0.1 is refusing connections. Anybody has any clue about this?

Regards
Kia

[phoenix]

Have you disabled the firewall and SElinux? You also need DNS A & MX records pointing to your server, are they?

[kibo]

Yes, SELinux is disabled on both Dom0 and guest and no firewalls are configured. The MX and A records are there and the dns is updated. Actually, I can see the e-mails coming in but they are not delivered to the local mail boxes. Also outgoing mails are not processed. I tried to test port 25 via telnet. It works fine and Postfix accepted the mail. But still it went to the deffered queue and stayed there!

Regards
Kia

[kibo]

I guess now I found a clue. In the defer folder of postfix, I opened one defered e-mail. It has following error in it:

reason=host 127.0.0.1[127.0.0.1] said: 451 4.5.0 Error in processing, id=05158-01, virus_scan FAILED: virus_scan: ALL VIRUS SCANNERS FAILED: ClamAV-clamd av-
scanner FAILED: Too many retries to talk to 127.0.0.1:3310 (Can't connect to INET socket 127.0.0.1:3310: Connection refused) at (eval 40) line 266. (in reply
to end of DATA command

I guess it is related to the antivirus. I did a telnet to port 3310, it opened it successfully and the clamav process is up. When I do zmcontrol start, it says OK to all the parts. But when I check in admin console in the server monitor section, it says the spam and av is down! very strange!

Now I have got another problem, apart from solving this. Because of this problem I had to repoint the dns back to the old server and start it. Now, I have around 150 e-mails in the defer folder of the postfix of this new server which I don't know how to transfer to the old one.

Any help would be appreciated
Regards

[tbullock]

Have you made any progress on this?

I am thinking about migrating from FC5 to RHEL, but I don't think I will do a complete tar of my current /opt/zimbra

I think I am going to pick and choose what I take over and I believe anything to do with clam and dspam I will leave behind. I think that last time I did this type of shuffle from FC3 to FC5, I did not take those directories as they appeared to not upgrade/migrate properly and there should really be nothing needed from them.

So you could try that. Migrating again, but leave the mods causing you grief out of the migration solution.

Just an idea with no gaurantees.

[klopfer]

I'm not sure how your IP addresses and DNS are setup on this virtualized server, but I had some similar issues getting zimbra installed on a RHEL 4 inside of vmware. Turned out I had two issues relating to DNS. I had a real IP address pointing to the server, but it came in through NAT on the host.

1) I needed to make sure that /etc/hosts properly listed localhost for 127.0.0.1 and the name of my server mapped to its internal IP address. This is described in the zimbra install docs.
2) I needed to get BIND running as described here
http://wiki.zimbra.com/index.php?title=Split_DNS

After that I needed to use the zimbra admin console to flush all of the messages once. Then it worked like a charm.