Linked by Eugenia Loli-Queru on Mon 21st Apr 2003 17:26 UTC
PostgreSQL is a robust, next-generation, Object-Relational DBMS (ORDBMS), derived from the Berkeley Postgres database management system. "PostgreSQL is one of the best-managed open source projects out there" a friend-in-the-know told me about 2 years ago, so today we feature a mini-interview with five members of the PostgreSQL team about their plans on the popular DB.
Permalink for comment
To read all comments associated with this story, please click here.
To read all comments associated with this story, please click here.
Recent Original Stories
Recent Comments
Headlines
Random Comments
Random Stories
Random OS Link
Well is pg_dump atomic AND non-blocking (alter tablespace <whatever> begin backup; is atomic and non-blocking)? If it isn't it's worthless as a true hot backup solution. I really don't like the idea of a global export tool being used as THE backup solution. You have to admit that using a sql export tool as the backup strategy is so 1980's.
You know I was going to say that you could just mine the log and find the sql statement and see if you can write a sql statement that could fix the bad update(s), but if your not in archivelog mode it's very possible that oracle overwrote the log that contained said sql statement so we are back to loosing all transactions sense your last backup. Nightly backups (hot or cold) are fine if you’re in archivelog mode. Just a little FYI, media failure is NOT the most common reason for a recovery.
Now would you run oracle without archiving the redo logs and using exp as your only backup solution? I sure wouldn't!
I would guess that 100% of your production oracle instances are in archivelog mode, but there are shops out there that do very dumb things every day. If not what do you do if someone makes an ad-hoc update (or more than one) with invalid data? Scrap all the transactions that have been made since your last backup? Now that sounds like a really progressive backup and recovery strategy
Seriously, why would you pay for an oracle license if your data wasn't important enough to run in archivelog mode? IMO you are using the wrong tool if this is acceptable.
BTW when I said incremental backups I wasn't thinking about a PITR I was thinking of how RMAN does incremental and cumulative backups. Yea and if anybody does make a RMAN work-a-like for postgresql for the love of god don't copy the user interface!
Also I know what the WAL stuff has been in PosgreSQL since 7.1 but it's incomplete IMO because this work can not be leveraged to do a PITR at this time.