Hi Readers,
I'm sponsoring a challenge to derive an accurate predictive model to predict when tables and indexes will benefit from a reorg:
http://dba-oracle.com/oracle_news/2005_2_1...rganization.htm In response, I got a fascinating response from a reader, with several great observations. To paraphrase:
- One thing that I have learned is that there are no "precepts" or "rules of thumb" that work in all situations.
- A global practice of re-organizing tables/indexes does not really help to accurately predict and detect performance bottlenecks.
- If I re-org my index for select, maybe my update/insert/delete will be much slower. You have to look to the application!
- I am still not sure that the "precepts" will generate anything really useful to the DBA's who are working in the trenches.
FYI, I’m working on a new book titled “Oracle Silver Bullets:
http://www.rampant-books.com/book_2005_1_silver_bullet.htm So, fellow DBA’s what is your experience?