Vmware with approx 1tb of VM's in a data center. I am using Veeam running in a windows VM, the backups to a local repo run very quickly.
I also want to offsite this using a "Backup Copy" job with WAN accelerators over a 100mb link to a low end xeon server using an external USB3 disk for the repo.
I have found this to be very slow, I appreciate the initial will be slow but I am worried the daily incremental will also be slow for a small amount of change (approx 2-5gb). From what I am reading the remote repo is synthetically created so you really need fast disk.
Is this correct? would I be better off just backing up direct to the remote repo and not using the backup copy? I am currently running the initial copy and dont want to wait 2 days for that to finish only to find the daily incremental copies take too long.
James, what is currently the bottleneck for the backup copy job? Your understanding is correct, it is recommended to have fast disks for for the backup copy job target, since WAN accelerators are effectively trading disk I/O for WAN bandwidth savings (add WAN cache to the equation). However, with a 100Mb link you probably don't need WAN accelerators as they were specifically designed for links slower than 100 Mbps. You could also test direct backup copy job mode and see whether your target storage is capable of transforming the amount of changed data your infrastructure produces in acceptable time.
