The following topics provide information about the mechanisms and options of data transfer from your workloads to their replicas.
A transfer method describes the way data is replicated from a source workload to a target. PlateSpin Forge provides different data transfer capabilities, which depend on the protected workload’s operating system.
For Windows workloads, PlateSpin Forge provides mechanisms to transfer workload volume data at either block or file level.
Windows Block-level Replication: Data is replicated at a volume’s block level. For this transfer method, PlateSpin Forge provides two mechanisms that differ by their continuity impact and performance. You can toggle between these mechanisms as required.
Replication using the Block-Based Component: This option uses a dedicated software component for block-level data transfer and leverages the Microsoft Volume Snapshot Service (VSS) with applications and services that support VSS. The installation of the component on your protected workload is automatic.
NOTE:Installation and uninstallation of the block-based component requires a reboot of your protected workload. No reboot is required when you are protecting Windows clusters with block-level data transfer. When you are configuring workload protection details, you can opt to install the component at a later time, deferring the required reboot until the time of the first replication.
Replication without the Block-Based Component: This option uses an internal ‘hashing’ mechanism in combination with Microsoft VSS to track changes on the protected volumes.
This option requires no reboot, but its performance is inferior to that of the block-based component.
Windows File-level Replication: Data is replicated on a file-by-file basis (Windows only).
For Linux workloads, PlateSpin Forge provides a mechanism to transfer workload volume data at block level only. Data transfer is powered by a block-level data transfer component that leverages LVM snapshots if available (this is the default and recommended option). See Knowledgebase Article 7005872.
The Linux block-based component included in your PlateSpin Forge distribution is precompiled for the standard, non-debug kernels of the supported Linux distributions. If you have a non-standard, customized, or newer kernel, you can rebuild the block-based component for your specific kernel. See Knowledgebase Article 7005873.
Deployment or removal of the component is transparent, has no continuity impact, and requires no intervention and no reboot.
The PlateSpin Server saves volume snapshots in the following directory by default:
\ProgramData\PlateSpin\Volume Snapshots
You might need to modify the path:
If the current drive for the path does not have sufficient space available for the Windows workload snapshots
If you want to move the location in order to more easily exclude the path from your backup list
You can use the PlateSpin Server configuration parameter VssSnapshotMountPath to specify a custom path on the server where you want to store the snapshots. If the parameter's value is empty, the path will remain at the default.
To specify a custom path for the volume snapshot directory on Windows:
Go to the PlateSpin Server configuration page at
https://<platespin-server-ip-address>/PlatespinConfiguration
Search for VssSnapshotMountPath, then click Edit.
In the Value field, specify the full path of the directory on the PlateSpin Server where you want to store volume snapshots for Windows workloads. For example:
G:\PlateSpin\Volume Snapshots
Click Save.
To make the transfer of workload data more secure, PlateSpin Forge enables you to encrypt data replication. When encryption is enabled, over-the-network data transfer from the source to the target is encrypted by using AES (Advanced Encryption Standard).
NOTE:Data encryption has a performance impact and might significantly slow down the data transfer rate by up to 30%.