A successful RMAN implementation will include the formal definition of both backup and recovery goals. You could refer back to Chapter 5 for RTO and RPO as a starting point. Mean Time to Recover (MTTR) is another objective that will be touched on in this chapter.
MTTR (also known as fast-start checkpointing) is not enabled in 11g by default. It allows the database to recover (apply committed transactions, rollback uncommitted), which automatically occurs while bringing up a database after a crash or during an actual restore and recovery session. To enable this feature, the following database parameter is set to a non-zero number.
fast_start_mttr_target integer 0
At the same time, reset the following database parameters back to zero:
LOG_CHECKPOINT_INTERVAL LOG_CHECKPOINT_TIMEOUT FAST_START_IO_TARGET
You should know from the previous chapters that faster checkpoints (synchronizing all the datafile headers, flushes data written to...