The speed of our VMWare is super slow. But we want to have backups in several places. We need duplicate job. I am curious - does duplicate job copy data from source or from backup set after which it is run? If it copy data from source it will be super slow, if it copy data from backup set (which reside on high speed storage) it will be super fast. I want that it copy data from backup set.
You mean from primary backup set? Not from source (VMWare)?
I am afraid that slow performance of backup is caused by speed of ESXi network itself and not much can be done here. We can of course allow BackupExec to access SAN directly, but if Windows will initialize all SANs the only solution for me and all my family will be to go and jump from the bridge.
Source --> ESX
Primary Backup ---> Backup from source to storage
Duplicate Set ---> Duplicate of the Primary Backup Set
backup via SAN, backing up the vmdks should be faster
Most customer use transport mode san for vmdk level backups and experience no problem.