Distribution Point Replication

I second this. I am currently using existing Windows servers for the SMB shares. I am leaning towards Robocopy scripts, however would also like a built-in way of doing this.

Robocopy is the way to go on Windows, afaik.

The only issue I see with creating a "built-in" mechanism is that there's a zillion different configurations one could use for serving up files. Now, who knows, maybe JAMF will make me look silly and come out with some Coded-All-Winter™ wicked solution next week.

I could at the minimum see JAMF providing a "best practice" recommendation for some of the major flavors of DP replication out there would be a great (and needed) start now that it seems like many folks are moving off of Apple hardware & software for their Casper infrastructure. It seems like we're all ripe to be re-inventing our own wheels.

Our JSS runs on Mac OS X server (10.6.8) and our DP's are all windows 2003/2008 shares. What I've setup is a launch daemon on our JSS that runs every 4 hours. When it launches it mounts the smb shares then updates them with rsync. We may move our JSS to a windows machine at some point and like Jared mentioned above, I would/will be using a scheduled task and robocopy to perform the updates.

We use a Carbon copy cloner scheduled task to handle the copy between our three distribution points twice a day. The master mounts the replicas and then copies data to them at 8 & 9 respectively. The real trick is making sure that your policies aren't going to execute before the packages have been synced between all the distribution points.

@nkoval I like the idea of using CCC. What settings have you setup for the replication from source to destination?
Im in the process of getting this setup and im concerned that if I modify files with the same name they will get missed in replication. Im interested in your feedback...