Oracle uses several structures to provide complete recovery from an instance or disk failure: the redo log, undo records, a control file, and database backups.
The Redo Log: is a set of files that protect altered database data in memory that has not been written to the datafiles. The redo log can consist of the online redo log and the archived redo log. The online redo log is a set of two or more online redo log files that record all changes made to the database, including uncommitted and committed changes. The background process LGWR writes the redo entries sequentially to an online redo log file. LGWR writes redo entries continually, and it also writes a commit record every time a user process commits a transaction.
Filled online redo files can be manually or automatically archived before being reused, creating archived redo logs. ARCHIVELOG and NOARCHIVELOG options.
Undo Records: undo records are stored in undo tablespaces. Oracle uses the undo data for a variety of purposes, including accessing before-images of blocks changed in uncommitted transactions. During database recovery, Oracle applies all changes recorded in the redo log and then uses undo information to roll back any uncommitted transactions.
Control files: The control files include information about the file structure of the database and the current log sequence number being written by LGWR. During normal recovery procedures, the information in a control file guides the automatic progression of the recovery operation.