1Z0-052 · Question #214
You are in the middle of a transaction and very crucial data has been modified. Because of a hardware failure, the instance has shut down before synchronizing all the database files. Which two…
The correct answer is A. On startup, SMON coordinates instance recovery. D. Uncommitted changes will be rolled back after the database is opened. After an instance crash, Oracle automatically performs instance recovery via SMON, which rolls forward committed changes using redo logs and then rolls back uncommitted changes using undo data once the database opens.
Question
You are in the middle of a transaction and very crucial data has been modified. Because of a hardware failure, the instance has shut down before synchronizing all the database files. Which two statements are true? (Choose two.)
Options
- AOn startup, SMON coordinates instance recovery.
- BOn startup, CKPT coordinates instance recovery.
- COn startup, use RMAN to perform instance recovery.
- DUncommitted changes will be rolled back after the database is opened.
- EOn startup, perform media recovery and then instance recovery.
- FOn startup, all the files will be synchronized and you get both committed and uncommitted data.
How the community answered
(41 responses)- A80% (33)
- B5% (2)
- C10% (4)
- E2% (1)
- F2% (1)
Why each option
After an instance crash, Oracle automatically performs instance recovery via SMON, which rolls forward committed changes using redo logs and then rolls back uncommitted changes using undo data once the database opens.
SMON (System Monitor) is the Oracle background process responsible for coordinating instance recovery on startup - it reads the redo logs to roll forward all changes since the last checkpoint and then initiates rollback of any uncommitted transactions, requiring no manual intervention.
CKPT (Checkpoint) updates control file and data file headers to record checkpoint SCNs, but it does not coordinate instance recovery; that responsibility belongs exclusively to SMON.
RMAN is a backup and recovery tool used for media recovery when data files are lost or corrupted; instance recovery after a crash is fully automatic and does not require RMAN.
After SMON completes the roll-forward phase and the database is opened, uncommitted transactions are rolled back using undo segments, ensuring that only committed data is permanently reflected in the database.
Media recovery is only required when data files are physically damaged or missing and must be restored from backup; a simple instance crash requires only automatic instance recovery, not media recovery.
Instance recovery does not preserve uncommitted data - after recovery the database will contain only committed changes, and any uncommitted modifications will be rolled back.
Concept tested: Oracle SMON-driven automatic instance recovery
Source: https://docs.oracle.com/en/database/oracle/oracle-database/19/cncpt/oracle-database-instance.html
Topics
Community Discussion
8
Prof. SaraMay 3, 2026A and D are your two correct picks. When Oracle detects a crash, SMON automatically performs instance recovery during startup by applying redo (roll forward) and then undoing uncommitted transactions, so you never touch RMAN for this scenario and you never manually trigger anything, the engine handles it entirely on its own.
Luis F.Apr 27, 2026SMON handles instance recovery automatically, then rolls back uncommitted txns.
Hiroshi T.Apr 29, 2026Worth adding that per the Oracle Database Concepts guide, instance recovery has two phases: cache recovery (SMON rolls forward using redo) comes first, and transaction recovery (rolling back uncommitted changes) can actually complete lazily after the database is already open to users, so SMON does not block availability until every rollback finishes.
Hiroshi T.May 1, 2026I almost picked F thinking sync meant clean data, but the Oracle docs confirm SMON handles instance recovery and undo rolls back uncommitted changes.
Ola B.Apr 15, 2026SMON handling instance recovery on startup is exactly right, and if you spin up a test database and kill it hard with an active transaction, you will watch SMON do its thing in the alert log during the next startup, rolling forward the redo and then rolling back anything uncommitted. D is also solid because that rollback of uncommitted changes is the second half of crash recovery, not a separate manual step you trigger. The trap in this question is C and E, since RMAN and media recovery are for restoring datafiles from backup after physical damage, not for the automatic crash recovery scenario described here. B is wrong because CKPT just writes checkpoint information, it does not coordinate recovery.
Hiroshi T.Apr 17, 2026Worth adding from the Oracle Database Concepts guide that the checkpoint position CKPT writes to the control file is exactly what SMON reads to determine where to begin rolling forward redo during instance recovery, so while B is wrong as a recovery coordinator, CKPT and SMON are more tightly coupled than the answer choices suggest.
Mateus R.Apr 12, 2026F tempts you, but SMON rolls back that mess. A, D.
Luis F.Apr 12, 2026Yeah SMON is the one doing cleanup on crashed instance recovery, though worth noting PMON is the one you think of first for process cleanup so it trips people up on the same question.