5 critical features of managed DB services @awscloud RDS PostgreSQL checks all five 👍
1/5⌚✅be able to recover to a point-in-time beyond the last backup (any point-in-time within the retention)
2/5⏱️✅enable automatic backups (and be warned if not). They occur during the defined window but do not block the database activity. And RDS takes an initial backup at creation.
3/5🦺✅ prevent accidental deletion of an instance (remember in case of sh.t happening, you are in stress and can make things worse)
4/5🪓✅do not delete the backups
when terminating the DB service (or be strongly warned) because backups also protects for accidental drop
5/5💾✅and at least keep a snapshot when you delete the instance and backups (snapshots are not backup, but it is ok for a terminated instance)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
0⃣ I've heard many misconceptions about restoring tablespaces in @OracleDatabase so here a a little #backToTheBasics restore+recover concepts & how-to
👇
1⃣ A tablespace is a subset of the (monolith) database. There are physical operations (backup, transport, snapshot,...) that you don't want to do for the whole TB database. Then you dedicate some tablespaces to applications, datastores, large tables, read-only workloads,...
2⃣You may want to restore some data as-of yesterday (because you dropped a table by mistake for example) without having to restore the whole database. This is Tablespace Point In time Recovery (TSPITR). Let's say you dropped the tablespace USERS by mistake at 11:42 👇