We are currently running an environment of a CASO server at our corporate office and MMS's at our remote locations. We are working on getting optimized deduplication running so that we can duplicate the backup sets from the remote MMS's to our CASO server over our WAN. A couple of questions I have regarding this configuration:
1. The "Private cloud server" setting (located on the "Settings" screen under the Storage tab in the console) for the CASO and each MMS. If the MMS are running optimized deduplication back to the CASO, should this setting be enabled on the CASO? What does enabling this setting actually? I have read through the Backup Exec 2012 Private Cloud Services Planning and Deployment Guide several times and it doesn't really explain what enabling that setting actually does.
2. That guide also mentions creating the following registry key:
HKLM\SOFTWARE\Symantec\Backup Exec for Windows\Backup Exec\Engine\Misc\DisablePDI2PDISetCopy
It says to do this "If you want to run incremental Exchange GRT duplicate backup jobs to the cloud, set the following registry value to 1 when the installation is complete. Changing the registry value disables the deduplication disk storage device's GRT-to-GRT duplicate copy capability on the Backup Exec server."
Is this the only case in which you should set this value (if you just want to run incremental Exchange GRT)? The reason I ask is that I also located the following support article http://www.symantec.com/docs/TECH172807 that states this value should be set to improve performance of optimized deduplication for GRT sets over a WAN, but since the article date is prior to BE 2012, I assume this is for BE 2010. Is this still true for BE 2012 and if so, is this required/recommended and should it be set on both the CASO and all MMS?
3. Does optimized deduplication over a WAN work well with backups run using the Hyper-V agent, or is it better to run the backups of the VM's individually using the Windows agents? I wasn’t sure if optimized deduplication was able to evaluate the contents of the VM’s in the Hyper-V backup and only get changed blocks instead of always duplicating the whole VM VHD file.
Any input on these questions would be appreciated.