Data flows drawn in the Protector UI are abstractions that hide the underlying hardware and software implementation. Before a data flow can be constructed, Protector Client software must be installed on the physical nodes connected to the storage devices, and the equivalent nodes must be created in the Nodes Inventory. The relationship between storage devices, Clients and nodes appearing in the Nodes Inventory is described in About nodes.
As an example of how a data flow is implemented, consider the backup of a file path on a server to a repository. In this case, the source and destination storage devices are file system volumes mounted on separate servers. Protector Client software must be installed on both servers. These Protector Clients are automatically detected by the Protector Master and appear in the Nodes Inventory as OS Hosts. The source OS Host can be used 'as-is' to identify the volume and file path in the backup policy. The Repository node is created via the UI; the destination OS Host node being selected as the proxy when configuring the node. When the rules for this data flow are activated, the source server will transmit the files identified in the policy over the network to the repository on the destination server. The figure below shows the data flow as it appears on the UI with the underlying implementation shown beneath.
A more complex example is a data flow representing application data, stored on a source block storage device, being replicated to a destination block storage device. The source block device has LDEVs mounted to an application server machine that are being used to store the application data. Protector Client software must be installed on the application server and on servers at the source and destination sites that are designated to control the block devices. The application server must have prerequisites installed to enable Protector to interact with the application software. The servers (ISMs) at the source and destination sites must have prerequisites installed to enable Protector to interact with the block devices. All three Protector Clients will be automatically detected by the Protector Master and appear in the Nodes Inventory as OS Hosts. Source and destination Block Device nodes are created via the UI; the source and destination OS Host nodes representing the ISMs being selected as proxies when configuring the nodes. When the rules for this data flow are activated, the source (VSP 1) will replicate the LDEVs identified in the policy over the data link to the destination (VSP 2). The figure below shows the data flow as it appears on the UI with the underlying implementation shown beneath. Note that the source Block Device node is required by Protector to control the replication, but does not appear on the data flow.