home.social

#oracledatabase — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #oracledatabase, aggregated by home.social.

  1. #OracleDatabase 23ai for on-premises #Exadata, offloading AI vector searches, and 17 microsecond latencies... OH MY. I am living in the future.
    blogs.oracle.com/exadata/post/

  2. Storage performance with #OracleDatabase on #Exadata can be mysterious, confounding, and outstanding especially when you have both DBRM and IORM active with multiple databases. Having a CDB-level Global AWR can give you all the deets on both layers to see what may and may not be happening. This new whitepaper from Oracle tells you more.
    oracle.com/docs/tech/database/

  3. Even on #Exadata, it will take a while to read 30+ TB from 3 tables. I think there is probably a way to optimize this query from an #OracleDatabase perspective. 🤨

  4. Really, Bard? At least tell me you're being aspirational or telling me what I want to hear. AFAIK there have never been nor can I find any reference to these #OracleDatabase parameters that could indirectly control behavior on #Exadata. 🙄

  5. Each day that goes by when I'm running queries on #OracleDatabase #Exadata, I hope that I will actually run into a case of "Delayed Block Cleanout" that I can confirm wrecks the performance of my query. 🤨

  6. Bloom filters on #OracleDatabase #Exadata can be offloaded IF the datasets are "small enough". What is "small enough"? And what about those ~10 hidden parameters that control bloom filters? Even my biggest Exadata customers can't get a clue from Oracle Development, and it's really ticking me off. 🤬