Apologies 6.5 hour period not 7.5. Basic Maths failing me.
The first thing to check is that the client machines have received the task and run it... check the agent log on the client machines: you're looking for something like "Invoking DAT Update Job".
Do you have entries like that on the offending machines?
Yeah the DAT job is setup correctly on all the machines I've managed to check. Over half of them so the chances are they all have it.
We rolled out the new SDAT (6006) yesterday and it seemed to propagate just fine, fairly quickly over the period of about 3 hours to all machines.
The biggest problem appears to be the reporting aspect more so than the DAT job updating now; most machines I've seen now have actually updated just fine, the console however still reports version 6006 instead of 6007, even though the machine is reporting 6007 from the shield in the systray.
I think ePO doesn't show the new DAT level until the workstation sends it's props. You might want to check your ASCI (agent to server communication interval).
ASCI is set to an hour for the clients in our dev environment.
Policy enforcement is every 5 minutes, comms to the management server is hourly.
It's very strange, left over night and I'm still showing 4 machines out of 38 that have reported back to the console, but every machine seems to have got the 6007 update.
Thanks for all the suggestions so far
It could have to do with this issue which I have seen manifest itself in HIPS patch reporting throughout our environment. This is from the pulled epo patch 2 readme (to be re-released as patch 3):
Issue: Summary product properties for managed systems such as version number, DAT version and Engine version may be incorrect and not match the detailed product properties which are correct. (Reference: 553913, 550837)
Resolution: The summary product properties now match the detailed product properties and are correct.
My earlier post on the issue: http://community.mcafee.com/thread/23456