Which statement is ture regarding checkpoints and recovery ?
-
A checkpoint defines the highest system change number (SCN)
-
All redo entries higher or at the SCN are known to be written to the data files.
-
Only the redo records containing SCNs higher then the checkpoint need to be applied during recovery.
-
The LOG_CHECKPOINT_INTERVAL initialization paramter specifies the amount of time beteen incremental checkpoints.
An Oracle checkpoint marks the highest system change number (SCN) guaranteed to be written to data files. During recovery, only redo records containing SCNs higher than the checkpoint need to be applied because all changes at or below the checkpoint are already safely on disk.
An Oracle checkpoint records the System Change Number (SCN) up to which all changed data blocks in the buffer cache have already been written to the data files by DBWn. This means during instance recovery, Oracle only needs to apply redo entries with SCNs higher than the checkpoint SCN, since everything up to and including that SCN is already durably on disk — matching the marked answer. The checkpoint doesn't define the 'highest' SCN in the sense implied by option 1 (it's the SCN guaranteeing durability, not necessarily the highest ever seen), option 2 is backwards (it's entries below/at the SCN that are already written, not 'higher'), and LOG_CHECKPOINT_INTERVAL specifies checkpoint frequency by OS blocks of redo, not by time interval.