Had a great first install of 7.2, this application is used in a testing scenario. Was happy to find that email transmission time was 1-2 seconds between the three virtual machines in my development test lab. This characteristic is very helpful for minimizing time overhead during repetitious test runs.
I have just moved all the VMs to the production site, which has one Windows Server 2008 virtualization host instead of the two Linux hosts I had developed on. Everything is working exactly as before, except that the Postfix message transmission time has increased by a factor of 10 to 20.
After comparing logs (/var/log/mail.log) I see "delay=0.75, delays=0.07/0.01/0/0.67" on the development Zimbra server, and "delay=10, delays=0.1/0/0/10" on the production Zimbra server.
Here is a diagram if that is helpful for clarity.
By the way, the production Server 2008 machine is very highly spec'd, with ample CPU and RAM, more than the development lab machines. Everything else in my automated test battery works perfectly. The email arrives, just too late. I would like to get the transmission delay close to what it is on my development equipment.
The delay seems to exposes a significant symptom of my problem. I would appreciate insights on where to look for the cause.