Relatively inexpensive solution with low data currency
Quickly roll back transactions if data corruption occurs
100% local protection
Staged backups and batch processing
Disaster tolerance, with varying degrees of data currency and performance impact
Snapshots are fairly inexpensive but it’s a point in time copy. Quick rollback from data corruption but no protection against primary logical unit failure.
Split mirrors – 100% copy of local data. Quick rollback from data corruption and protection against primary logical unit failure. But we can’t hold as many splits as we can snapshots because it takes up the same amount of disk capacity as the original volume.
Neither of these protect against subsystem failure or site failure. So here we probably want to use it in conjunction with remote mirrors
A lot of the time we are probably going to use in-system replication and remote replication at the same time so we will have multiple copies of our data.
This is all about operational recovery. We’ve also got backups which is our last line of defence.