ReferenceGlossary
Point-in-time recovery
Restoring a database to its exact state at a chosen moment, by loading a base backup and then replaying the write-ahead log up to that time. Because the log is archived continuously, you lose only the last few minutes of writes, not a whole day.
Data
In Quorum#
The Point-in-time recovery Research adds Backups to the Database card: None, Nightly, or PITR. Nightly snapshots can lose a calendar day of writes and add 5% to the Database's Burn. PITR loses about 2 calendar minutes and adds 15%. The DR drill for the IPO checklist reads this RPO and passes only under 5 minutes, unless a replica in another region already covers the data.
Read next#
RPO and RTO, Replication lag, Read replica, Encryption at rest