You can choose to perform latest recovery or a point in time recovery based on your requirements.
A latest recovery operation only maintains the most recent copy of the data at the destination volume.
Point in time Recovery
A point in time recovery operation maintains multiple recovery points at the destination volume.
Additional options appear when Point in time recovery is selected:
Recovery point store: Select the Recovery Point Store on the destination server where journal entries for each recovery point are stored. To configure a RP store, see Configuring the RP store for File System.
Crash-consistent recovery points interval: Select this option to create crash-consistent recovery points, and then specify the time interval at which recovery points are created.
Application-consistent recovery points interval: Select this option to create application-consistent recovery points, and then specify the time interval at which recovery points are created.
When this option is selected, source volumes are quiesced to create application-consistent recovery points for the destination volumes.
Retain recovery points for: Select this option to specify how long journal entries for a recovery point should be retained.
After this time period, the accumulated change history for recovery points is pruned from the RP Store and written to the destination volumes.
Merge recovery points older than: Select this option to combine older recovery points into larger recovery points that span a longer time range (for example, daily instead of hourly), and then specify the number of days that more frequent recovery points are retained.
Recovery point interval at the end of retention: Select this option to specify the time interval between the oldest recovery points in the RP Store.
As recovery points get older and are merged, the interval between them grows larger. This option configures the maximum allowed interval for the oldest recovery points.
Switch to latest Recovery if RP Store is offline for more than: Use this option to switch to live replication if the RP Store on the destination server is inaccessible for the time interval you specify.
If you specify this option, the replication service abandons the RP Store if it does not come back online within the specified period. The replication service resyncs the destination volume with the source volume, and switches to live replication instead of granular replication. When this option is used, and the RP Store comes back online after the specified interval, all changes in the RP Store are erased and all accumulated recovery points in the RP Store are discarded.
If you do not specify this option, the replication is paused, and the change journal accumulates on the source server until the RP Store comes back online.
With this option, you can choose whether to a) preserve the changes in the RP Store, but not apply them to the destination server while the RP Store is offline; or b) continually update the destination server, but lose all granular recovery points.
Prune and merge changes during off-peak time only: Use this option to specify whether to transfer updates for the oldest recovery points to destination server during off-peak hours.
If you specify this option, recovery point updates are transferred to the RP Store only during off-peak hours, providing better replication performance during peak hours.
If you do not specify this option, recovery point updates are transferred to the destination server on an ongoing basis, based on the value specified for the Change journal retention interval option.
With this option, you can prioritize a) replication speed or b) reclaiming space in the RP Store.