I would certainly not personally abandon traditional backup methods completely, but look to create a mixed policy which still utilised tape for point-in-time backups and the replication function of the SAN to handle incremental backup tasks. This way you can effectively trash your incremental backup snapshots once you have verified your full tape job and begin the process again – this should ensure the DR SAN does not become overloaded with obsolete snapshots.
Also, if you then experience data corruption which is replicated, or total hardware loss (eg, in a fire) it will be possible to do a full restore from your tapes, therefore giving an effective worst-case scenario get-out-of-jail-free card.
Related Q&A from Martin Taylor
Before choosing a RAID configuration for a new storage-area network (SAN), you need to consider the performance requirements of your data and the ...continue reading