Moved Zimbra VM now cannot send/receive

[killc]

We have been testing Zimbra Network Edition in a VMWare Server 2.0 environment and it has been working fine (send, receive, archiving, etc.) As a test we stopped the VM, copied all of the files to a new server and ran the VM on the new host. We brought down the old server and configured the host and guest with the same IP addresses and brought it all up.

There was an issue where Ubuntu Server 8.04 (the guest OS) was not seeing eth0 on the copied guest, but that seems a common problem and while it doesn't see eth0, it does see eth1 so I configured that with the correct IP address.

Zimbra started up fine from zimbra user (zmcontrol start).

I could log in as admin and a user via another computer on the Internet and everything is reported as working in the admin panel. All of the users and data seems to be present as it was before.

I attempted to send an email from the new Zimbra VM and it appeared to send but it was never received by the external email address. I tried sending mail into the new Zimbra server from the external address and it never appeared either. I then telnetted to port 25 from a computer on the Internet successfully and attempted to send an email... all of the commands went through fine and it said the message was queued but it also did not get sent.

I am stumped as to what could be the disconnect here, and any help would be great!

Thanks!

[uxbod]

Sounds like a DNS issue to me so please post the following

cat /etc/hosts
cat /etc/resolv.conf
dig yourdomain mx
dig yourdomain any
host `hostname` <- note backticks and not single quotes

[killc]

It appears that the issue is with VMWare and the fact that Linux (Ubuntu) binds the eth# to a MAC address. When the VM was moved, a new fake MAC was created and since it didn't jive with the original one it sees not eth0 and now sees only eth1. I am working to fix this but if anyone has any experience here or knows of an easy solution, do tell.

[veronica]

does mail stay in queue ?
what does /var/log/zimbra.log says when you send mails ?

[bdial]

you probably have to remove the old mapping from /etc/udev/rules.d/

[Klug]

Originally Posted by killc

It appears that the issue is with VMWare and the fact that Linux (Ubuntu) binds the eth# to a MAC address. When the VM was moved, a new fake MAC was created and since it didn't jive with the original one it sees not eth0 and now sees only eth1. I am working to fix this but if anyone has any experience here or knows of an easy solution, do tell.

Remove (or fix) the "hwaddress" line in /etc/network/interfaces.