The replication mode is "Synchronous". This means that when the host server commits a write to a SAN disk, that bit of data will be written to the DR array in Norwest first, the DR array then sends an acknowledgement back to the production array and that bit of data will be written to the prod array when it receives that acknowledgement from the DR side. Meaning that the data between the R1 (prod) and R2 (DR) devices are always in-sync when the SRDF mirrors are established.
• When the SRDF mirror/replication is established, the process write-disables the target devices on the DR array in Norwest, so your DR server will not be able to access the DR SAN devices while that is on. If you want to make the DR LUNs available to the DR host, we will have to split the SRDF mirror for that set of servers.
• Zero data loss? As long as the data has been committed to the SAN disks, it should be covered. As for data that might still be sitting in the server cache but have NOT been flushed out/committed to SAN disk, well, the SAN storage has no control over data that has not been written to it.