Cloud Connect Backup Copy

[mongie]

Hello All,

Just wanting to confirm my understanding of how a Backup Copy job works when using Cloud Connect.

Should I expect to write only incremental backups over the Internet/WAN connection, and then have a the GFS transform happen between the cloud gateway and cloud storage?


Thanks,
Alex

[foggy]

Yes, only changes are sent to the cloud repository. GFS transform happens on the repository server (or between gateway server and the storage, in case of CIFS repository).

[veremin]

However, cloud gateway server is not taking part in this process. Thanks.

[dellock6]

I've seen some confusion in this topic because some service providers are using the cloud gateways also as repositories, but this is just a corner case where the two roles are installed in the same machine, but the involved role is still the repository and no other one.

[mongie]

so... what "node" is performing the transform on the cloud hosted backups?

I'm trying to establish what sort of bandwidth would be required for a particular set of data (50GB Change, 3TB Full).

[dellock6]

As said, every transform operation is done by the target repository at the service provider. If daily change is 50GB, this is the amount of data that will be sent each day over the internet, regardless the type of backup copy (regular or GFS), minus the possible data reduction if wan accelerator is used.