Oracle DBA · Lesson 13B

Oracle DBA Lesson 13B — Find an Archive Destination Problem

A data-file backup captures one point in time. Archived redo can extend recovery from that backup toward a later committed state. This lesson traces how completed redo reaches an archive destination — and how to diagnose what happens when a required destination or recovery area cannot accept more files.

Watch on YouTube

Oracle DBA Lesson 13B — Find an Archive Destination Problem

AI Ops Pro

Archived redo extends the recovery span

An archiver process (ARCn) copies a completed online redo group to configured archive destinations. Those copies support media recovery choices after file loss — they are how you roll a restored backup forward through later committed work.

Instance recovery is different. After an instance failure it uses online redo. NOARCHIVELOG still supports instance recovery, but its media-recovery options are severely limited. ARCHIVELOG is about keeping a durable trail beyond the online groups.

When a destination problem matters

A destination problem matters when the configured archive requirement cannot be satisfied. Not every destination is necessarily mandatory. Oracle evaluates destination attributes (mandatory / optional) and the configured minimum-success count. If a required copy cannot complete, an old online redo group may remain unavailable for reuse — and log switching can eventually stall.

Diagnose the actual requirement. Do not assume that any one failed destination automatically stops the database.

Three questions: mode, health, capacity

Start with database log mode. Then read each active archive destination’s status, destination path, and exact error. Finally, inspect recovery-area capacity when an FRA is configured. These views answer three separate questions: mode, destination health, and capacity.

Log mode:

SELECT log_mode
FROM   v$database;

Destination health (status · destination · error):

SELECT dest_id, status, destination, error
FROM   v$archive_dest_status
WHERE  status <> 'INACTIVE';

FRA capacity (when configured):

SELECT name, space_limit, space_used, space_reclaimable
FROM   v$recovery_file_dest;

Read absence, reclaimability, and errors carefully

• No FRA row can mean no FRA is configured — that is not automatically a fault.
• Space reclaimable depends on retention and deletion policy. It is not permission to remove files manually with an OS command.
• A destination error directs investigation toward storage, connectivity, or configuration. Preserve the exact error text and timestamps.

Relieve pressure without sacrificing recovery

Protect recoverability while relieving pressure. Practical options include:

• Adding real capacity
• Correcting a failed destination
• Completing the required backup or offload
• Performing retention-aware RMAN cleanup under an approved policy

Never delete archive logs with an operating-system command as a quick space fix. Enabling ARCHIVELOG also does not replace a tested backup — recovery still needs usable backups, the needed redo, and a tested procedure.

Gotchas

• Not every failed destination stalls reuse — configured requirements decide.
• Instance recovery ≠ media recovery; archives matter for the media path.
• Empty v$recovery_file_dest often means no FRA, not a broken FRA.
• Reclaimable space ≠ license to rm archive files by hand.
• Keep exact ERROR strings and timestamps from v$archive_dest_status.

Quick quiz

Choose one answer, then submit. You’ll see the correct answer and a short why.

1. Why does ARCHIVELOG mode matter for recovery choices after file loss?

It replaces the need for any data-file backup Archived redo can extend recovery from a backup toward a later committed state It turns off instance recovery so LGWR never writes online redo It makes every optional destination mandatory automatically

2. A destination shows an error. Does that always stop log reuse?

Yes — any error on any destination freezes the database immediately Yes — only if the destination path contains FRA No — only when the configured archive requirement can no longer be met No — destination errors never affect reuse

3. Which trio of views answers mode, destination health, and FRA capacity?

v$database, v$archive_dest_status, v$recovery_file_dest v$session, v$process, v$locked_object v$datafile, v$tempfile, v$controlfile v$log, v$logfile, v$parameter only

4. SPACE_RECLAIMABLE looks large under pressure. What should you not do?

Add real capacity if the FRA is undersized Correct the failed destination path or connectivity Run retention-aware RMAN cleanup under an approved policy Delete archive logs with an OS command as a quick space fix

AIOps Pro

https://www.apps-dba.com/ https://x.com/AIOpsPro https://www.youtube.com/@AIOpsPro