1Z0-908 flashcards
1Z0-908 · Build recall with a focused review session.
34 cards. Read the term and the definition here. The deck at the top of this page is optional.
- SGA components
- The System Global Area contains the database buffer cache, shared pool, large pool, redo log buffer, and fixed SGA, sized automatically when SGA_TARGET is set.
- SPFILE vs PFILE
- The SPFILE is a binary server parameter file changed with ALTER SYSTEM SCOPE; the PFILE is a text file edited manually and read at startup or with STARTUP PFILE.
- Database startup stages
- NOMOUNT starts the instance and reads the parameter file; MOUNT reads control files; OPEN opens data files and redo logs. Recovery may stop at MOUNT.
- ARCHIVELOG mode requirement
- ARCHIVELOG mode must be enabled for online backups and point-in-time recovery. Switching requires the database in MOUNT state and a restart of the instance.
- Datapatch purpose
- After OPatch applies binaries to an Oracle Home, datapatch runs the corresponding SQL patches against each database, updating the registry views.
- Release Update patching
- Release Updates and Monthly Recommended Patches replace the older PSU/CPU model; they are applied with OPatch to the Oracle Home, then datapatch to each database.
- Read-only Oracle Home
- A read-only Oracle Home shares an unpatched base with multiple patched homes, reducing disk use and patch time. ORACLE_HOME and ORACLE_BASE must not be nested.
- Locally managed tablespace
- A locally managed tablespace stores its extent bitmap inside the tablespace rather than in the data dictionary, avoiding recursive space transactions and contention.
- ASSM and PCTFREE
- Automatic Segment Space Management uses bitmaps in segment headers for free space and requires PCTFREE for row growth but ignores PCTUSED, which applies only to manual management.
- Bigfile tablespace
- A bigfile tablespace contains exactly one data file and can reach about 32 petabytes with 8K blocks; it uses a 4G extent bitmap and simplifies file management.
- Undo vs temp tablespace
- Undo stores rollback and read-consistency data for transactions; TEMP stores sort and hash work areas and temporary segments. Both are mandatory in every database.
- WITH ADMIN OPTION
- Granting a system privilege or role WITH ADMIN OPTION lets the grantee grant it onward. Revoking the original grant does not cascade to dependent grantees.
- WITH GRANT OPTION
- Granting an object privilege WITH GRANT OPTION lets the grantee grant that privilege on the object. Revoking it cascades to privileges granted from it.
- Profile password limits
- Profiles control FAILED_LOGIN_ATTEMPTS, PASSWORD_LIFE_TIME, PASSWORD_REUSE_MAX, and PASSWORD_VERIFY_FUNCTION, plus resource limits such as SESSIONS_PER_USER and IDLE_TIME.
- SYSDBA versus SYSOPER
- SYSDBA can start up, shut down, back up, recover, and create or drop a database. SYSOPER can start up, shut down, back up, and recover but cannot create or drop.
- RMAN incremental levels
- Level 0 is the base backup. Level 1 differential backs up blocks changed since the last level 0 or 1; cumulative backs up blocks changed since the last level 0.
- Fast Recovery Area sizing
- DB_RECOVERY_FILE_DEST sets the FRA location and DB_RECOVERY_FILE_DEST_SIZE sets its limit. Space pressure causes obsolete backups and archived logs to be deleted.
- OPEN RESETLOGS
- Incomplete recovery ends with OPEN RESETLOGS, which creates a new database incarnation and resets the redo log sequence, invalidating prior archived logs.
- Flashback Database
- Flashback Database rewinds the whole database to a past point using flashback logs, requires MOUNT state, and needs DB_FLASHBACK_RETENTION_TARGET and a configured FRA.
- AWR snapshot interval
- AWR captures performance statistics every 60 minutes by default; DBMS_WORKLOAD_REPOSITORY.MODIFY_SNAPSHOT_SETTINGS changes the interval and retention period.
- DB time formula
- DB time equals CPU time consumed by sessions plus non-idle wait time. It excludes idle wait events and measures total database workload elapsed time.
- ASH sampling
- Active Session History samples active sessions every second into V$ACTIVE_SESSION_HISTORY, letting you see which wait events and SQL consumed time.
- SQL Plan Management
- SPM stores accepted execution plans as baselines so the optimizer uses only verified plans, preventing performance regressions after statistics or parameter changes.
- Oracle isolation levels
- Oracle supports READ COMMITTED, SERIALIZABLE, and READ ONLY. It never permits dirty reads, and readers do not block writers or vice versa.
- ORA-01555 cause
- Snapshot too old occurs when undo information needed for read consistency has been overwritten, typically from long queries or frequent commits in loops.
- Row lock contention wait
- A session waiting for another session to commit or roll back a locked row reports the enq: TX - row lock contention wait event in V$SESSION.
- Deadlock detection
- Oracle detects deadlocks automatically after about three seconds, rolls back one statement with ORA-00060, and writes a deadlock trace to the alert log.
- CDB and PDB structure
- A container database holds CDB$ROOT, PDB$SEED, and one or more pluggable databases. Each PDB has its own data dictionary and open mode.
- Common versus local users
- Common users are created in CDB$ROOT with a C## prefix and can connect to any container. Local users exist in a single PDB only.
- Local undo in CDB
- LOCAL_UNDO_ENABLED gives each PDB its own undo tablespace, allowing PDB cloning and point-in-time recovery without affecting the CDB root.
- Dynamic service registration
- PMON registers the instance and its service names with the listener using LOCAL_LISTENER. The database must be at least in NOMOUNT state for registration.
- EZConnect syntax
- EZConnect connects without client files using host:port/service_name, for example host01:1521/orclpdb, and is enabled by the ezconnect naming method.
- DBMS_SCHEDULER job
- A scheduler job combines an action, a schedule or repeat interval, and a job class. DBA_SCHEDULER_JOB_RUN_DETAILS reports each run's outcome.
- Data Pump REMAP_SCHEMA
- Data Pump import with REMAP_SCHEMA=old:new loads objects into a different schema, and REMAP_TABLESPACE redirects storage to another tablespace.