daily 18:00 UTC · 30d retention · transaction-log archive · managed bucket
Backups & PITR
Physical backup with hot and cold modes. The PITR window is the recovery window you've configured.
Recovery objectives
given current config
RPO · max data loss
—
Tighten by increasing backup frequency or enabling continuous log streaming.
RTO · target restore time
~ —
Restore as new cluster (default). In-place is faster but destructive.
PITR · recovery window
pick a target LSN or wall-clock
—now
selected target
— · LSN —
No transaction-log archive for this instance yet. Point-in-time recovery replays sealed log segments shipped to object storage — once continuous archiving has run, a point-in-time window appears here. Until then, recovery granularity is the daily snapshot interval.
| when | type | size | target | status | retention | verified |
|---|---|---|---|---|---|---|
| loading… | ||||||
Size = the provisioned volume size, not bytes stored. EBS snapshots are incremental — each keeps only the blocks changed since the previous one, so storage cost is far below the figure shown.
point-in-time restore · destructive
Restore to selected point
This rewinds the database to (LSN ). Writes after that point are discarded. Restore runs as a new cluster by default; in-place is irreversible.
Type restore to point to confirm. Self-serve execute is operator-assisted today — this opens the request with your exact target.