The hardware replication acts exclusively at the physical layer of the I/O stack. It offers advantages such as speed and platform independence. However, because this replication occurs at a low level in the I/O stack, the real work must be done at the OS level to fully access the replicated data. This can be challenging and many IT infrastructures consider a simple approach: removing the LVM layer. Using this approach, the major complexity of accessing the data disappears. However, to have performance file systems (as needed for advanced applications, databases, SAP...), system administrators require the use of LVM, partly because the LUN management system embedded in storage arrays is not sufficient for I/O intensive applications. The method described in this article shows it is possible to take advantage of a LVM in the hardware replication infrastructure.