How Carbonite Recover works

Begin with servers you want to protect to the cloud. These servers are referred to as source servers.

To protect your source servers, you must have at least one target appliance for Windows, one for Linux, and at least one worker. Carbonite will create a worker for you, but you must create the target appliance.

• Target appliance—A target appliance is a virtual server in the cloud that was created from a template provided by Carbonite. You must have at least one target appliance for Windows and one for Linux, and they can protect multiple source servers. However, you may need additional target appliances if you are protecting a larger number of disks or to help balance the load when protecting many servers. The target appliance maintains a replica of the data from the source servers you are protecting, and in the event of a failure, the data on the target appliance is used to quickly failover to a replica virtual machine in the cloud.
• Workers—Workers receive and execute tasks by communicating with the Carbonite Recover backend infrastructure that Carbonite is running.

When protection begins, the target appliance maintains a replica of the data from the source servers you are protecting by using virtual hard disks attached to the target appliance.

In the event a source server fails, the worker quickly creates a replica virtual machine in the cloud and detaches the hard disks from the target appliance and attaches them to the new replica virtual machine.

You can run on the replica virtual machine in the cloud as long as needed. When you are ready, you can restore and failback from the replica virtual machine in the cloud back to your original server or to a different server, as needed.